Skip to main content

Full text of "An interpretation of Rumphius's Herbarium amboinense"

See other formats


AN INTERPRETATION OF 
RUMPHIUS'S HERBARIUM 
AMBOINENSE 



BY 

E. D. MERRILL 




144971 



MANILA 
BUREAU OF PRINTING 
1917 



A. S. H.APR 2* W» 



PUBLICATIONS FOR SALE BY THE BUREAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS 



ETHNOLOGY 

A VOCABULARY OF THE IGOROT LAN- 
GUAGE AS SPOKEN BY THE 
BONTOC IGOROTS 

By Walter Clayton Clapp 

Order No. 408. Paper, 89 pages, $0.75, 
postpaid. 

The vocabulary is given in Igorot-English 
and English-lgdrot. 



THE NABALOI DIALECT 

By Otto Scheereb 
and 

THE BATAKS OF PALAWAN 

By Edward Y. Miller 

Order No. 403. Paper, $0.25; half mo- 
rocco, $0.75; postpaid. 
The Nabaloi Dialect (65 pages* 29 
plates) and the Bataks of Palawan (7 
pages, 6 plates) are bound under one cover. 



THE BATAN DIALECT AS A MEMBER 
OF THE PHILIPPINE GROUP 
OF LANGUAGES 

By Otto Scheerer 
and 

"F" AND "V" IN PHILIPPINE 
LANGUAGES 

By Carlos Everett Conant 

Order No. 407. 

These two papers are issued under one 
cover, 141 pages, paper,' $0.80, postpaid. 



THE SUBANUNS OF SINDANGAN BAY 

By Emerson B. Christie 

Order No. 410. Paper, 121 pages, 1 
map, 29 plates, $1.25, postpaid. 

Sindangan Bay is situated on the north- 
ern coast of Zamboanga Peninsula. The Su- 
banuns of this region were studied by Mr. 
Christie during two periods of five and six 
weeks, respectively. 

The 29 plates illustrate the Subanuns at 
work and at play; their industries, houses, 
altars, and implements; and the people 
themselves. 



THE HISTORY OF SULU 

By Najeeb M. Saleeby 

Order No. 406. Paper, 275 pages, 4 
maps, 2 diagrams, $0.75, postpaid. 
In the preparation ef his manuscript for 
The History of Sulu, Doctor Saleeby spent 
much time and effort in gaining access 
to documents in the possession of the Sultan 
of Sulu. This book is a history of the 
Moros in the Philippines from the earliest 
times to the American occupation. 



ETHNOLOGY— Continued 

STUDIES IN MORO HISTORY, LAW, 
AND RELIGION 

By Najeeb M. Saleeby . 

Order No. 405. Paper, 107 pages, 16 
plates, 5 diagrams, $0.25; half mo- 
rocco, $0.75; postpaid. 

This volume deals with the earliest 
written records of the Moros in Mindanao. 
The names of the rulers of Magindanao are 
recorded in five folding diagrams. 



NEGRITOS OF ZAMBALES 

By William Allan Reed 

Order No. 402. Paper, 83 pages, 62 
plates, $0.25; half morocco, $0.75; 
postpaid. p 

Plates from photographs, many of which 
were taken for this publication, show orna- 
ments, houses, men making fire with bamboo, 
bows and arrows, dances, and various types 
of the people themselves. 



INDUSTRIES 
PHILIPPINE HATS 

By C. B. Robinson 

Order No. 415. Paper, 66 pages, 8 
plates, $0.50 postpaid. 

This paper is a concise record of the 
history and present condition of hat making 
in the Philippine Islands. 



THE SUGAR INDUSTRY IN THE 
ISLAND OF NEGROS 

By Herbert S. Walker 

Order No. 412. Paper, 145 pages, 10 
plates, 1 map, $1.25, postpaid. 

Considered from the viewpoint of prac- 
tical utility, Mr. Walker's Sugar Industry 
in the Island of Negros is one of the most 
important papers published by the Bureau 
of Science. This volume is a real contribu- 
tion to the subject; it is not a mere com- 
pilation, for the author was in the field and 
understands the conditions of which he 
writes. 



A MANUAL OF PHILIPPINE SILK 
CULTURE 

By Charles S. Banks 

Order No. 413. Paper, 53 pages, 20 
plates, $0.75, postpaid. 

In A Manual of Philippine Silk Culture 
are presented the results of several years' 
actual work with silk-producing larvae to- 
gether with a description of the new Philip- 
pine race. 



1a ak interpretation of 
^rumfhTus's herbarium 
amboinense - 



BY 

E. D. MERRILL 




144971 



MANILA 
BUREAU OF PRINTING 
1917 



Department of Agriculture and Natural Resources 
BUREAU OF SCIENCE 
Manila 

Publication No. 9 



(Actual date of publication, November 1, 1917.) 



Dedicated to the Memory 

OF 

Charles Budd Robinson, Jr. 
Pictou, Nova Scotia, October 26, 1871 
t Amboina, December 5, 1913 

3 



CONTENTS 



Par*. 



Illustrations 7 

Preface 9 

Introduction 11 

The importance of the Herbarium Amboinense 11 

Rumphius and his work ~ 15 

Amboina 18 

The genesis of the plan and the botanical exploration of Amboina 20 

Errors in the interpretation of Rumphian species 27 

The interpretation of Rumphian species as types 29 

The interpretation of the species described in the Herbarium 

Amboinense by various authors 31 

Linnaeus and Stickman 31 

Burman 34 

Buchanan-Hamilton 35 

Henschel 36 

Hasskarl 36 

The present status of Rumphian species 38 

Difficulties in interpreting Rumphian species 40 

Nomenclature 43 

Overlooked binomials 46 

Acknowledgments 50 

Systematic Enumeration 53 

Thallophyta „ 53 

Pteridophyta 63 

Spermatophyta 74 

Gymnospermae 74 

Angiospermae 79 

Monocotyledonae 79 

Dicotyledonae 179 

Species described or mentioned by Rumphius that cannot be 

definitely referred to their proper families 504 

Sequence of Species in Rumphius's Herbarium Amboinense with 

their Binomial Equivalents _ 511 

Addenda 549 

Index 551 

5 



ILLUSTRATIONS 



Plate I. Map, showing Amboina and the surrounding islands. 

II. Map of Amboina. 
Text figure 1. Form of field label. 



PREFACE 



It was with considerable diffidence and great regret that I 
assumed the task of interpreting the species described in the 
Herbarium Amboinense, as it was at my suggestion that Doctor 
Robinson undertook this task, in the prosecution of which he 
met his untimely death. 

Doctor Robinson arrived in Amboina July 15, 1913, and 
immediately commenced his field work, which was actively 
prosecuted up to the time of his death. On the morning of 
December 5, 1913, he departed from the town of Amboina, un- 
accompanied, for a botanical excursion through the country south 
of the town, passing through Amahoesoe, Eri, Silalei, Latoelahat, 
and Aerlo. While on his return trip he was murdered by the 
Boetonese residents of a small settlement between Aerlo and 
Seri, about 15 kilometers from the town of Amboina. The crime 
was wholly due to a local superstition, the ignorant natives 
mistaking Doctor Robinson for the notorious potong kapala 
(the decapitator) , who is currently believed to wander about 
during November and December for the purpose of cutting off 
human heads.* 

When Doctor Robinson left Manila for Amboina in June, 
1913, he fully expected to complete the task of interpreting the 
species described in the Herbarium Amboinense, but he lived 
to finish only a part of the necessary field work. His un- 
foreseen death, while he was actively prosecuting his field work, 
placed an entirely different aspect on the problem, and in order 
that his untimely death might not have been in vain, I felt 
constrained to take up the task, at the point where Doctor 
Robinson's labors were ended, and to carry the project to 
completion, so far as this could be done from the material and 
data available. 

The qualifications necessary to obtain the best results in this 
special field are exacting. A wide knowledge of the Malayan 
flora is essential, as is a thorough understanding of Latin, of 
Dutch, and of the native names of plants in the Malay Archi- 
pelago, qualifications to which I can lay claim only to a limited 

* Merrill, E. D. Charles Budd Robinson, Jr. Philip. Joum. Sci. 9 (1914) 
Bot. 191-197. 

9 



10 



RUMPHIUS'S HERBARIUM AMBOINENSE 



degree. Under the circumstances the future investigator will 
doubtless find in the present work errors both of omission and 
of commission. Much remains to be done before all the species 
that are typified by the Rumphian descriptions are thoroughly 
understood, but a great part of this work will depend on a 
continued and comprehensive botanical exploration of Amboina 
and of the neighboring islands with special attention to this 
object. 

The actual working up of the results could best have been 
done at one of the large European botanical institutions, where 
access could be had to large libraries, to comprehensive collec- 
tions of botanical material from all parts of the world, and to 
type and authentically named specimens, as well as the oppor- 
tunity of consulting specialists in various groups. As a trip to 
Europe was impracticable, the work was done in Manila, utilizing 
the local library and herbarium. 

In the prosecution of the task I have been obliged to interpret 
numerous species from their published descriptions, as herbarium 
material representing them was not available in Manila. Like- 
wise, there are a number of botanical works containing refer- 
ences to the Herbarium Amboinense that I have been unable to 
consult, as no copies of these were to be had in Manila. This 
difficulty has been overcome in part by borrowing certain es- 
sential works, and in part by sending to various botanical 
institutions for copies of original descriptions, which have gener- 
ously been supplied by botanists in the United States and in 
Europe. In general, while it is realized that the present inter- 
pretation of the species described in the Herbarium Amboinense 
is incomplete and imperfect, it is also realized that completeness 
and perfection in this difficult task are relative terms and that 
many of the species that I have been obliged to enumerate as 
of more or less doubtful status could not have been more de- 
finitely placed, even if I had had access to all of the botanical 
literature and an opportunity to examine all of the extant 
botanical material from the Indo-Malayan region. It is felt, 
however, that the present treatment of the Rumphian species 
meets a real requirement and that, to a very large degree, it 
will clear the way for the more intensive study of the problems 
in connection with each individual species of doubtful status. 

E. D. Merrill. 



INTRODUCTION 



THE IMPORTANCE OF THE HERBARIUM AMBOINENSE 

The Herbarium Amboinense consists of twelve books, published 
in six volumes. The purpose of each book is expressly stated, 
thus: 

Liber primus. Qui continet arbores, quae f ructus esculentos f erunt, 

ac culturam humanam requirunt. 
Liber secundus. Continens arbores aromaticus, quae aut fructum 

aromaticum, corticemve, aut odoratum aliquod praebent lignum. 
Liber tertius. Continens arbores, quae Resinam, speciososque dant 

Flores, aut noxium aliquod lac fundunt. 
Liber quartus. Arbores continens silvestres, quarum pars aliqua fab- 

rilis est. 

Liber quintus. De Arboribus agens silvestribus promisque. 
Liber sextus. De fruticibus agens tam domesticis, quam silvestribus. 
Liber septimus. De funibus agens silvestribus & Fruticibus reptan- 
tibus. 

Liber octavus. De plantis agens domesticis, tam victui, quam Medi- 

cinae, & decori inservientibus. 
Liber nonus. De Convolvulis, & Herbis reptantibus. 
Liber decimus. De Herbis agens silvestribus promiscue. 
Liber undecimus. Agens de reliquis herbis silvestribus. 
Liber duodecimus. De arbusculis agens marinis, & plantis saxosis, 

seu de Lithodendris & Lithophytis. 

To the above should be added the "Auctuarium," cited in this 
work as volume seven of the Herbarium Amboinense, which 
contains additional notes on species described in the above 
twelve books as well as figures and descriptions of plants not 
included in them. 

The classification is primarily the ancient one of trees, shrubs, 
and herbs, with subdivisions according to habitats and uses. 
There is no system based on other than the most evident, gross 
characters. While the purpose of each book is definitely in- 
dicated, it is important to understand the object of the work 
as a whole. Rumphius expressly states that it was based, not 
on Amboina alone, but on all of the Dutch East Indies. How- 
ever, as the work was done on plants or parts of plants either 
growing in Amboina, sent to him from other regions, or brought 
in from other countries for commercial purposes, he selected the 
title Herbarium Amboinense. Numerous species now growing 

11 



12 



RUMPHIUS'S HERBARIUM AMBOINENSE 



in Amboina are not described in the Herbarium Amboinense, 
for Robinson's collection alone presents more than 350 species 
of ferns and flowering plants not considered by Rumphius.* 
Some of these have undoubtedly been introduced into Amboina 
since Rumphius's time, but very many of them are indigenous 
and were certainly as common there in the seventeenth century 
as they are to-day. Some of the species not appearing in the 
Herbarium Amboinense are small and might have been ignored 
as being of slight importance ; but others are large trees, shrubs, 
or vines, often with rather showy flowers, and in many cases 
they are abundant. Doctor Robinson's final conclusions as to 
the methods used by Rumphius in selecting the species described 
are expressed in a letter written early in November, 1913, 
as follows : 

I think more and more that the Herbarium Amboinense was not at all 
a complete flora of Amboina as Rumphius found it and that he selected 
on four bases: Economic plants and others that resembled them; plants 
that were very different from those he had seen in Europe; plants that 
greatly resembled those of Europe; plants regarding which there was 
some superstition or legend. A fifth heading might be made for the very 
showy plants, but I think that this really belongs under the second group. 

The Herbarium Amboinense is a classical work on the Malayan 
flora, and one that is absolutely essential to the systematist 
to-day. This is not because of any system of classification pro- 
posed, for the work follows no definite system, nor on account of 
priority of its names, as the work is pre-Linnean, and binomial 
names appearing in the text are merely accidental. The great 
importance of the work is due to the fact that later authors have 
made the Rumphian descriptions and figures the actual "types" 
of many binomials. As an original source the Herbarium Am- 
boinense stands preeminent among all the early publications on 
Malayan botany. In more than 800 original "publications" of 
species of plants under the binomial system from 1753 to 
1908 the Rumphian names or figures, or both, are quoted as 
synonyms, and in about 350 cases the proposed binomials are 
based wholly on data given by Rumphius. In no case is a 
species typified by Rumphian figures and descriptions intelligible 
without reference to the Herbarium Amboinense. These nu- 
merous species not represented by any "type" specimens must, of 
course, be interpreted primarily by the data given by Rumphius, 
supplemented by a study of botanical specimens from the same 

* Merrill, E. D. Reliquiae Robinsonianae. Philip. Journ. Sci. 11 (1916) 
Bot. 243-319. 



INTRODUCTION 



13 



general region from which Rumphius secured his material. In 
spite of what has been accomplished in the past hundred years on 
the Moluccan flora and the intensive field work prosecuted in Am- 
boina for four and one-half months by Doctor Robinson, num- 
erous species typified by Rumphius's descriptions and figures 
are still of doubtful status and must so remain, until in each 
case they are definitely connected with botanical material orig- 
inating in the classical locality for each species and agreeing 
with the descriptions and figures, the native names, the economic 
uses, and the other characters indicated by Rumphius. 

The work already prosecuted in Amboina and the neighboring 
islands has yielded material by which the essential characters 
of very many of the Rumphian species can be definitely deter- 
mined, but much remains to be done in this field. In botanical 
literature there are scores of species whose only published de- 
scriptions are the brief general statements compiled from the 
Herbarium Amboinense, from which data alone it is usually im- 
possible for the working systematist to gain any definite idea 
of the true characters of the species. This is especially true 
in such critical genera as Calamus, Elaeocarpus, Citrus, Bambusa, 
Canarium, and in many others. Botanists generally have been 
content to work on the Malayan flora, describing as new the 
various forms that have appeared in current collections, without 
making any serious attempt to determine the exact status of 
species in the same groups based on Rumphian descriptions. 
Stability in nomenclature demands that the status of these early 
species be determined as soon as possible, for otherwise many 
reductions must be eventually made. 

The Herbarium Amboinense was very extensively cited by Lin- 
naeus's contemporaries and successors, especially by Burman f., 
Loureiro, and others, who wrote on the floras of regions geo- 
graphically allied to Amboina, and by all authors of general 
works on systematic botany up to the middle of the nineteenth 
century. In the more recent works on systematic botany the 
Herbarium Amboinense is not so frequently quoted as in older 
ones, references to this work being to a large degree those 
necessary to explain synonymy. However, binomials based 
wholly on the Rumphian descriptions and figures continue to be 
proposed, the latest ones observed being Sindora galedupa Prain, 
1897, and Calamus acidus Beccari, 1906. 

It is by no means certain that the importance of the Her- 
barium Amboinense is fully appreciated. The number and the 
size of the volumes, seven, folio; the number of printed pages, 



14 



RUMPHIUS'S HERBARIUM AMBOINENSE 



over 1,660; and the number of plates, about 695, give but an 
inadequate idea of the immense amount of data contained 
in this work. Rumphius described in greater or less detail and 
named about 1,700 forms. His descriptions, while sometimes 
scarcely more than casual mention, are more often very ample, 
and to these are often appended discussions of the economic uses 
of the various plants described. Few works on Indo-Malayan 
botany, published since the Herbarium Amboinense was written, 
can compare with it in amount and variety of original data. 
When it is fully realized that practically all of the immense 
mass of data included in the Herbarium Amboinense represents 
the observations and accumulated knowledge of one man, the 
great energy and ability displayed by Rumphius in preparing this 
monumental work, which was accomplished under very adverse 
circumstances, can be more fully appreciated. 

The work is immensely more than a discussion of the plants 
of Amboina. While it is true that most of the descriptions and 
the greater part of the figures were based on Amboina specimens, 
copious references are found to other regions, extending from 
Madagascar to China and Japan, southeast to New Guinea, and 
even to Mexico and South America. It is very evident that 
Rumphius's colleagues and correspondents transmitted material 
to him from the whole Orient, and he incorporated his descrip- 
tions of this material in the Herbarium Amboinense. 

Among Doctor Robinson's papers were found the following 
compiled data, which will give some definite idea of the regions 
covered by the work. The most important references to Java 
comprise 125 entries; to Celebes, 83; to Ceram, 77; to Bali, 74; 
to Banda, 53; to Buru, 42; to the Moluccas proper, including 
Ternate, Tidore, Batchian, and Halmaheira or Gilolo, 58; to the 
Philippines, 20; to Boeton, 12; to Borneo, 8; to Sumatra, 8; 
to Madura, 4; to Manipa, 15; to the Sumbawa-Timor group, 
including Sumbawa, Timor, Nussa Radja, Solor, Wetter, and 
Rottea, 23 ; to the Aru Islands, 6 ; to the Key Islands, 3 ; to New 
Guinea, 4; to the Sula Islands, 8; and to the small islands near 
Amboina, 8. There are numerous references to China, but fewer 
to Japan, Indo-China, Malacca, Madagascar, southern Africa, 
Mexico, Peru, and Brazil. A number of the species figured and 
supplied with ample descriptions were based on this extra- 
Amboina material, although others are only casually mentioned. 
It is at once evident that by no means all the Rumphian species 
can be interpreted from Amboina material and data alone. 



INTRODUCTION 



15 



RUMPHIUS AND HIS WORK 

George Everhard Rumphius, as the family name Rumpf or 
Rumph is Latinized, well named "the Pliny of the Indies," was 
born in 1627, apparently in Hanau, Hesse Cassel, Germany, and 
died in Amboina, June 15, 1702, at the age of 75 years. Detailed 
accounts of his life and work are available in the writings of 
numerous authors* so that it is unnecessary to enumerate here 
more than the most important facts in connection with the 
preparation and publication of his most renowned work, the 
Herbarium Amboinense. 

Rumphius entered the service of the Dutch East India Com- 
pany as a young man, proceeded to Batavia, Java, in 1653, and 
in the latter part of the same year to Amboina, where he resided 
for the remainder of his life. Perhaps for the first two years 
of his stay in Amboina he was stationed at Larike, but later he 
was transferred to Hila. It is evident that he commenced the 
preparation of the Herbarium Amboinense shortly after his 
arrival in Amboina, his active work being continued practically 
until his death, in spite of the great handicap of blindness after 
the year 1670. In 1670, while still stationed at Hila, he had the 
work about completed, and it was then his intention to return 
to Europe. To make a more definite study and comparison of 
all his species, he undertook a final series of journeys along 
the coasts and in the hills, and to this he himself attributes his 
blindness which followed almost at once. His published works 
are manifestly based largely on observations made by him 
between 1653 and 1670. The handicap of blindness was some- 
what lessened by aid given him by his wife and by assistants 
assigned to him by the Dutch East India Company. In 1673 
he removed to the town of Amboina and commenced to translate 
the original Latin text of the Herbarium Amboinense into Dutch. 
In the following year, however, his wife and eldest child were 
killed in the great earthquake of that year, and subsequent to 
that date he had less other assistance, some of it of little real 
value. The original illustrations for the Herbarium Amboin- 
ense were apparently made by Rumphius himself, but on Jan- 
uary 11, 1687, Amboina was visited by a disastrous fire, in 
which Rumphius's house was destroyed, including his library, 
many of his manuscripts, and the plates of the Herbarium Am- 

* See Rouffaer and Muller "Biographien van Rumphius" in their: 
Eerste proeve van een Rumphius-Bibliographie. Rumphius Gedenkboek 
1702-1902 (1902) 176-185. 



16 



RUMPHIUS'S HERBARIUM AMBOINENSE 



boinense. Undaunted by this last catastrophe, he replaced the 
destroyed illustrations by new drawings, some made by his son, 
P. A. Rumphius, others made by various assistants supplied 
by the East India Company. Thus in attempting to interpret 
Rumphian species the fact must be constantly kept in mind that 
the illustrations were not made from the actual specimens on 
which the corresponding descriptions were based. In this con- 
nection I venture to give the following translation of Rumphius's 
own statement : * 

The plates were drawn, by various artists, some of the figures larger, 
some smaller, but each marked with its name, which I myself never saw 
[italics mine] ; but I have learned from the skilled, and am informed that 
they sufficiently agree with the plants themselves, but what ought some- 
times to be changed I have marked on the plates themselves or in the text. 
However, the reader may take these as sufficiently faithful and pleasing, 
while perchance they may be corrected by others, or better ones produced, 
for he will readily perceive that in this country I have not been provided 
with the best artists, for which reason also I have not been ashamed to 
refer him to other works with larger and better plates, especially Rheede, 
which has recently been published. 

This passage explains much in connection with the Herbarium 
Amboinense, such as the very crude execution of some figures and 
the excellent reproduction of others; the union, in a few cases, 
of the characters of totally different species on a single drawing 
such as Pemphis and Aegiceras, and Urena and Triumfetta; the 
fact that certain drawings do not conform to the characters 
given in the description that they are supposed to represent ; the 
absence of drawings to illustrate species that are fully described ; 
why certain species, certainly as common in Amboina in Rumph- 
ius's time as they are to-day, are not mentioned in the work; 
and perhaps the rather striking discrepancies in the magnifica- 
tions or reductions of parts of various species so evident on 
many of the drawings. 

In 1690 the manuscript of the first six books was delivered to 
the Dutch East India Company, the remaining parts being de- 
livered in 1695. The manuscript of the first six books was 
forwarded to Holland from Batavia, Java, in 1692 on the Water- 
land. This ship was destroyed by the French in transit, and the 
manuscript was lost with the ship. Fortunately a copy had been 
retained, and thus the fruit of Rumphius's many years of labor 
was not lost. A copy of these six books was finally sent to 
Holland in 1696, the manuscript of the remaining six books was 



* Herb. Amb. 1 (1741) author's preface, last page. 



INTRODUCTION 



17 



sent the following year. The manuscript of the "Auctuarium," 
completed by Rumphius in 1701, a few months before his death, 
was copied at Batavia and sent to Holland in 1704. This im r 
portant manuscript remained in the archives of the Dutch East 
India Company until 1736, when the company granted permis- 
sion to Professor J. Burman to prepare it for printing, the six 
volumes appearing between 1741 and 1750 and volume seven, 
the "Auctuarium," in 1755. 

The general title of the published work, taken from volume 
three, is as follows: 

Georgii Everhardi Rumphii, | Med. Doct. Hanavensis, Mercatoris Sen- 
ioris, & in Amboina Consulis, nomine | Plinii Indici Celebris, & Membri 
Inlustris Societatis Aca- | demiae Naturae Curiosum Germaniae, | Herba- 
rium | Amboinense, | Plurimas complectens Arbores, Frutices, Herbas, Plan- 
tas terrestres & aquaticas, j quae in Amboina, | et ajdacentibus reperiuntur 
insulis, | Adcuratissime descriptas juxa earum formas, cum diversis denom- 
inationibus, | cultura, usu, ac virtutibus. | Quod & insuper exhibit | varia 
insectorum animaliumque genera, | Plurima cum naturalibus eorum figuris 
depicta. | Omnia magno labore ac studio multos per annos conlecta, | & 
duodecem conscripta libris. | Nunc primum in lucem edita, & in Latinum 
sermonem versa, | Cura & Studio | Joannis Burmanni, | Med. Doct. et in 
Horto Medico Amstelaedamensi Professoris | Botanici, Academiae Caesareae 
Naturae Curiosum Socii; | Qui varia adjecit Synonyma, suasque Observa- 
tiones.* 

Volume seven, the Auctuarium, issued five years after volume 
six was printed, bears the following title page : 

Georgii Everhardi Rumphii | * * * | Herbarii | Amboinensis | 
Auctuarium, | Reliquas complectens Arbores, Frutices, ac Plantas, | quae 
in Amboina, et adjacentibus demum repertae sunt insulis, | Omnes accu- 
ratissime descriptae, & delineatae juxta earum | formas, cum diversis In- 
dicis denominationibus, | cultura, usu, ac viribus; | Nunc primum in lucem 
editum, & in Latine sermonem versum, | Cura & Studio | Joannis Bur- 
manni |***| Q u j V aria adjecit Synonyma, suasque Observationes.f 

In the present paper the Auctuarium has been consistently 
cited as volume seven of the Herbarium Amboinense. 

While Rumphlus's fame rests largely on the Herbarium Am- 
boinense, this by no means represents all that he accomplished. 

* 1 (1741)* 1-200, t. 1-82, preface, introduction, etc.; 2 (1741) 1-270, 
t. 1-87; 3 (1743) 1-218, t. 1-U1; 4 (1743) 1-154, t. 1-82; 5 (1847) 1-492, 
t. 1-181* ; 6 (1750) 1-256, t. 1-90. The title page varies somewhat in the 
different volumes and between editions one and two, a second edition, not 
essentially different from the first, having been issued in 1750. The Dutch 
title page is not here repeated. 

t(1755) 1-74, t. 1-29, Index Universalis [1-20]. The Dutch title page 
is not here repeated. 

144971 2 



18 



RUMPHIUS'S HERBARIUM AMBOINENSE 



His Rariteitkamer * is in itself a remarkable book, in which 
are figured and described numerous crustaceans, echinoderms, 
starfishes, several hundred mollusks, both univalves and bivalves, 
crystals, fossils, and other forms. This work passed through 
several editions, and the figures are in general distinctly better 
than those in the Herbarium Amboinense. His Amboinische 
Historie and Amboinische Land-Beschrijving still remain in 
manuscript. His Amboinische Dierboek was planned, accord- 
ing to Leupe, to consist of three books, in which the birds, the 
land animals, and the marine animals were to be described and 
figured, this to supply for the animal kingdom what the 
Kruidboek (that is, the Herbarium Amboinense) did for the 
plant kingdom. This work, however, was never published as 
such, although Valentijn apparently utilized much of Rumphius's 
data, perhaps not always acknowledging its source. His manu- 
script reports on the agriculture of Amboina, on the fortifications 
of Castle Victoria, description of the Amboina earthquake, a 
Malay dictionary, and other writings are still extant. Rumph- 
- ius's activities as an investigator other than as a student of 
plants are here briefly mentioned merely to emphasize the ability, 
energy, and broad interest of the man, for his record as an inves- 
tigator is a most remarkable one, more especially so when we 
take into consideration the period in which he lived and worked 
and the great handicaps under which he struggled. 

AMBOINA 

Amboina, Amboyna, or Ambon, as the name is variously 
spelled, is a small island situated about 128° east and 4° 40' 
south, a short distance south of the western end of Ceram on 
the north side of Banda Sea and not far from the western end 
of New Guinea. In the history of Malayan botany it is of 
preeminent importance, as it is the type locality of several 
hundred species, many of which were very imperfectly character- 

* D' Amboinsche | Rariteitkamer, | Behelzende eene beschryvinge van 
allerhande | zoo weeke als harde | Schaalvisschen, | te weeten raare | 
Krabben, Kreeften, | en diergelyke Zeedieren, | als mede allerhande | 
Hoorntjes en Schulpen, | die men in d' Amboinsche Zee vindt: | Daar bene- 
vens zommige | Mineraalen, Gesteenten, | en soorten van Aarde, die in d' 
Amboinsche, en zom- | mige omleggende Eilenden govonden worden. | Ver- 
deelt in drie Boeken, | En met nodige Printverbeeldingen, alle naar 't 
leven getekent, voorzien. | Beschreven door | Georgius Everhardus Rum- 
phius, | van Hanauw, Koopman en Raad in Amboina, mitsgaders Lid van 
het Kyzerlyke kweekschool der | onderzookers van de Natuurkuunde in 't 
Duitsche Roomsche Ryk opgerecht onder den naam van | Plinius Indicus | 
(1705) XXVIII + 1-340 [43], t. 1-60. 



INTRODUCTION 



19 



ized by the early authors. From a botanical standpoint it 
owes its great importance almost wholly to the preparation and 
publication of Rumphius's Herbarium Amboinense. (See Plates 
I and II.) 

During the early colonial period Amboina was of great com- 
mercial importance on account of the dominance of the spice 
trade, of which it was the center for a long time. It was first 
visited by the Portuguese in 1511, who established a factory 
there in 1521. The Portuguese were dispossessed by the Dutch 
in 1609, who have since retained control of the island except 
for the periods 1796-1802 and 1810-1814, when it was occupied 
by the British, being finally restored to the Dutch Government 
in 1814. The island is only 51 kilometers long and has an area 
of approximately 950 square kilometers. Salahoetoe, the highest 
mountain on the island, attains an altitude of 1,027 meters. 

The flora of Amboina is typically Malayan, although a few 
Australian types are present as in other parts of the Malayan 
region. Practically all of the species found along the seacoast 
are of general distribution from India to Malaya and Polynesia, 
Likewise most of the species found in the settled areas at low 
and medium altitudes, weeds of cultivation, and the generally 
cultivated economic and ornamental plants are the same as those 
usually found throughout Malaya, very many of which are now 
distributed in all tropical regions. The primeval forest to a 
large extent has been destroyed at low and medium altitudes, 
at least in those regions best adapted to agricultural pursuits, and 
has been replaced over large areas by cultivated or semicultivated 
plants, second-growth forests, thickets, and open grasslands 
characterized by the dominance of the cogon or lalang grass 
(Imperata) . In the interior on the slopes of the higher moun- 
tains, such as Salahoetoe, some forest still persists. 

The island presents a considerable endemic element, but a 
thorough botanical exploration of the Moluccas will doubtless 
show that most of the species now known only from Amboina 
inhabit also the neighboring islands, such as Ceram, Buru, 
Boeton, Celebes, Gilolo, western New Guinea, and the numerous 
smaller islands of the Moluccas. From the standpoint of en- 
demic species most of the neighboring islands are probably of 
much greater interest than is Amboina, but from the standpoint 
of the history of Malayan botany, no part of the Moluccas can 
be compared with it. 

No description of the vegetation of Amboina is here attempted, 
as I have not personally visited the island, and Doctor Robinson 



20 



RUMPHIUS'S HERBARIUM AMBOINENSE 



left no notes regarding the general character and appearance of 
the plant life of Amboina. However, a brief, general description 
has been given by Karsten.* The general appearance of the 
vegetation is apparently the same as that of similar regions in 
the Malay Archipelago and the Philippines, not subject to a 
prolonged dry season, where the original vegetation has not 
been entirely destroyed by man. 

THE GENESIS OF THE PLAN AND THE BOTANICAL EXPLORATION OF 

AMBOINA 

The logical and simple plan of exploring Amboina with the 
special object of collecting and studying the Rumphian species 
in their native habitat in connection with all data given by 
Rumphius, while perhaps conceived by other botanists, has previ- 
ously been carried into effect only by the late Dr. J. G. Boerlage 
of the botanic garden at Buitenzorg. In 1900 Doctor Boerlage, 
accompanied by Dr. J. J. Smith, made a trip to Amboina for the 
explicit purpose of collecting in the classical localities the plants 
described by Rumphius, more especially material represent- 
ing those species on which binomials of later authors had 
been based. Most unfortunately Doctor Boerlage contracted a 
fever while in Amboina, which resulted in his death at Ternate, 
August 25, 1900, while returning to Java, with the consequence 
that the results of his field work were never made available. 
Unquestionably, many botanists who have visited Amboina 
and carried on field work there have realized that it was a clas- 
sical locality in Malayan botany and that botanical specimens 
from that island would be of special value in interpreting Rumph- 
ian species, yet no single large collection has ever been made 
in Amboina of which the duplicates were given a wide distribu- 
tion, so that the general results of previous botanical work in 
Amboina have not been available to many botanists who have 
had occasion to discuss Rumphian species. While the present 
consideration of the species described and figured in the Her- 
barium Amboinense is of necessity incomplete, and doubtless 
errors in interpretation have been committed both in reference 
to Rumphian species and to binomials, yet it is felt that the 
work, somewhat in the nature of an innovation in systematic 
botany, is a step in advance and that it should prove to be merely 
preliminary to more intensive field work in relation to the same 

* Morphologische und biologische Untersuchungen iiber einige Epiphy- 
tenformen der Molukken. Ann. Jard. Bot. Buitenzorg 12 (1895) 117-195, 
t. 18-19. 



INTRODUCTION 



21 



general problems not only in reference to the Herbarium 
Amboinense, but also to other pre-Linnean works of similar 
importance. 

Certain post-Linnean works are susceptible of the same general 
treatment, especially those, like Blanco's Flora de Filipinas, 
in which the various species described are not represented by 
extant botanical material or types, but must be interpreted solely 
by the descriptions and data given by the author. In this con- 
nection I have in the past four years made an intensive study 
of all the Philippine species described by Blanco and have 
prepared for distribution to the larger botanical institutions of 
the world an extensive exsiccata which I have called "Species 
Blancoanae." The specimens selected for this exsiccata are 
those which I have determined to represent the Blancoan species, 
and to a large degree these specimens will take the place of 
Blanco's types, none of which were preserved by him. My 
Philippine experience in attempting to interpret Blancoan species 
logically lead to the application of the same general methods in 
reference to those figured and described by Rumphius. 

In 1902, on commencing botanical work in the Philippines, 
I was immediately confronted with the problem of interpreting 
the numerous forms described by Blanco in his Flora de Fili- 
pinas,* totaling about 1,130 species and varieties, of which not 
a single one is represented by type material, for Blanco preserved 
no herbarium specimens. Blanco's species, often very imper- 
fectly described and frequently placed in the wrong genus, have 
for the most part not been clearly understood by subsequent 
authors and as a result very many of them appear in botanical 
literature as doubtful or imperfectly known ones. The average 
botanist, working in Europe or America from dried specimens 
alone, with few or no field notes and with no personal knowledge 
of the Philippines and its vegetation, has found it impossible 
properly to interpret very many of Blanco's species. The 
clue to the identity of a Blancoan species is frequently found, 
not in the description itself, but in the appended economic data, 
native names, and other information given by Blanco. 

My experience in the interpretation of Blanco's species con- 
vinced me that the same methods, if applied to Amboina in 
connection with all the data given by Rumphius in the Herbarium 
Amboinense, would certainly yield material by which a high 
percentage of the several hundred species proposed by various 

* Blanco, M. Flora de Filipinas (1837) LXXVIII + 1-887 ; ed. 2 (1845) 
LXIX + 1-619. 



22 



RUMPHIUS'S HERBARIUM AMBOINENSE 



authors from 1753 to date might be elucidated. Rumphius's 
species, like those of Blanco, are not represented by extant 
botanical material, although it is manifest that Rumphius pre- 
served, at least temporarily, specimens representing some of the 
forms that he described.* Thus, in interpreting Rumphian 
species, we must utilize the same general methods as those 
devised in securing material and data to locate Blanco's species. 
In the interpretation of many Rumphian species the investigator 
has the great advantage of the published figures, but it should 
be carefully noted that the figures do not always correspond 
with the characters of the species indicated in the descriptions 
(see p. 41). In interpreting Blancoan species there is the 
distinct advantage of his use of technical terms and the rather 
doubtful advantage of the binomial nomenclature; doubtful not 
because of the system, but from Blanco's erroneous interpreta- 
tions of so many genera. 

At various times, as it became necessary to interpret species 
in critical genera by consultation of the Herbarium Amboinense, 
the necessity of securing botanical material from Amboina be- 
came apparent. The desirability of securing Amboina material 
was discussed by Doctor Robinson and myself at various times 
during his first tour of duty in the Philippines, 1908-1911. As 
work on the Philippine flora progressed, the necessity for a 
definite knowledge of the Rumphian species became more and 
more evident, until finally the matter was taken up with the 
Philippine authorities, approval of the Amboina project secured, 
and cooperation with the authorities of the botanic garden at 
Buitenzorg, Java, arranged. 

In preparation for his work in Amboina, Doctor Robinson 
made an exhaustive examination of the Herbarium Amboinense, 
and compiled on index cards all data that might be of assistance 
in his actual field work. He arranged all native names cited 
by Rumphius, for he realized that the clue to the identity of 
a Rumphian species would often be found in the native name 
or names cited. Regions and localities from which Rumphius 
secured his material were also classified, so that before com- 
mencing field work in Amboina, Doctor Robinson could deter- 
mine to a great extent what species he might hope to find in 
Amboina and what would have to be sought for in other islands. 

* Martelli, U. Le collezioni di Giorgio Everardo Rumpf acquistate dal 
Granduca Cosimo III de'Medici, una volta esistenti nel Museu di Fisica e 
Storia Naturale di Firenze, estratto da un catalogo manoscritto dal Prof. 
Giovanni Targioni-Tozzetti (1903) 1-213. 



INTRODUCTION 



23 



A special field label was prepared for the work, on which he 
recorded data of value in connection with the problem of the 
determination of Rumphian species and the results of his pre- 

FLORA OF THE MALAY ARCHIPELAGO 

HERBARIUM, BUREAU OF SCIENCE, MANILA, P. I. 



Common name Dialect 

Field No Herbarium No 

Collector, C. B. Robinson. 

Island 

Locality 

Habitat 

Altitude above the sea m. 

Tree; shrub; woody vine; herbaceous vine; herb 

Height of plant m; diameter. cm. 

Flower 

Fruit 

Supposed to represent 



Rumph. Herb. Amb. 



Identification considered certain; probable; possible; very doubtful. 



Date , 1913. 

Fig. 1. Form of field label. 

liminary comparison of the actual specimens with the Rumphian 
figures and descriptions. This field label is shown in fig. 1. 

His material, as collected, was compared with the Rumphian 
descriptions, and his conclusions were noted on the field labels. 
Specimens that were certainly, probably, or possibly identical 



24 



RUMPHIUS'S HERBARIUM AMBOINENSE 



with forms that Rumphius considered were placed in one series, 
and those that could not be referred to Rumphian species were 
placed in another. As material was matched with forms named 
and described by Rumphius, such species were checked on a 
special list. The two series established by Doctor Robinson in 
the field have been the basis of the two series into which the 
collections were finally divided for purposes of study. The 
specimens that could certainly or with a fair degree of cer- 
tainty be referred to Rumphian species were placed in the 
series "Plantae Rumphianae Amboinenses," thus cited in the 
present work, while the remainder were placed in the series 
' 'Reliquiae Robinsonianae" and are the basis of a separate re- 
port.* Of the Plantae Rumphianae Amboinenses, the labels of 
which bear both the Rumphian name and reference and the 
binomial as determined by the accepted code of nomenclature, 
there are about 600 numbers; of the Reliquiae Robinsonianae, 
including the cellular cryptogams, there are about 960 numbers. 
In arranging this material and in its critical study, a few speci- 
mens have been transferred from one series to the other. In 
both series collections made at different dates and with separate 
field numbers have been combined when presenting the same 
stage of development and unquestionably representing the same 
form. The data compiled by Doctor Robinson, as a result of his 
field observations and the comparison of the fresh material with 
the Rumphian descriptions, has been of immense value in the 
preparation of the present work. 

During the prosecution of his field work in Amboina, it became 
evident to Doctor Robinson that he could not expect to find all 
the forms figured and described by Rumphius, nor even all of 
those that were from Amboina. In his progress report, written 
from day to day, he frequently mentioned the slow progress 
of the work and his disappointment in not being able to locate 
this or that species. He commenced his field work with the 
idea of taking specimens only from plants found in flower or 
in fruit, but he occasionally collected single specimens from 
sterile plants for purpose of check. It is evident that he had 
located a number of species described by Rumphius of which he 
collected no botanical material, but which he was watching in 
the hope that he could later find them in flower or in fruit. At 
various times he indicated his purpose to collect sterile material 
of the species that he could not find in flower or in fruit before 

* Merrill, E. D. Reliquiae Robinsonianae. Philip. Journ. Sci. 11 (1916) 
Bot. 243-319. 



INTRODUCTION 



25 



his final departure from Amboina, a plan that was never carried 
out on account of his sudden and unexpected death. Many of the 
very common species, such as the coconut, the betle nut palm, the 
papaya, and numerous cultivated ornamentals, are lacking in 
the collection chiefly for the reason that the actual preparation 
of specimens of these common and well-known plants was pur- 
posely deferred until the more important and critical species 
had been secured. 

In the prosecution of his field work in Amboina, Doctor Rob- 
inson was handicapped by the same factors that have hindered 
our attempts to secure material in the Philippines to clear up 
the status of Blanco's species. With the increase of population in 
Amboina, as in the Philippines, the original vegetation has been 
totally destroyed over large areas, the virgin forest being replaced 
by grasslands, thickets, and second-growth forests of a type 
entirely different from the original vegetation. Many species 
definitely mentioned by Rumphius as occurring in specific local- 
ities can no longer be found in the indicated places. It is by 
no means improbable that many species, common in Amboina in 
the seventeenth century, have now become extinct there, or at 
least are very rare and local, even as various Philippine species 
mentioned by Blanco as occurring in definite localities can no 
longer be found within many miles of the respective places 
mentioned by him. Native names given by Rumphius have in 
many cases become obsolete or are so altered as to be hardly 
recognizable, although in many cases the name cited by 
Rumphius is still in use and for the same species under which it 
was cited by him. Rumphius, like Blanco, secured most of his 
material from the settled areas and from the forests at low 
altitudes, and it is unfortunately true that, in the Malayan region, 
the forest vegetation at low altitudes is the vegetation most 
rapidly destroyed by the encroachment of man. 

The practical extermination of the original vegetation of those 
regions best adapted to agricultural pursuits is a subject that 
deserves more consideration than it has received.* Unquestion- 
ably, many species of plants have been exterminated in various 
parts of the Malayan region within the past century as the 
population has increased. The areas devoted to agriculture are 
rapidly being enlarged in many parts of this vast region, and the 
consequent destruction of primeval forests over large areas is 
a strong argument in favor of a vigorous and intensive botanical 

* Merrill, E. D. Notes on the flora of Manila with special reference to 
the introduced element. Philip. Journ. Sci. 7 (1912) Bot. 145-208. 



26 



RUMPHIUS'S HERBARIUM AMBOINENSE 



exploration of Malaya, in order that representatives of certain 
elements of the flora shall be secured while they are still available 
or at least easily accessible. A continued and intensive explora- 
tion of the Moluccas is greatly to be desired before the actual 
plants that will yield material to clear up various Rumphian 
species of doubtful status shall have become extinct or at best 
local and of rare occurrence. 

It was originally planned that Doctor Robinson should pro- 
secute his field work for about four months, but as the work 
progressed it became increasingly evident to him that this period 
of time was altogether too short. On the basis of representations 
made by him, Doctor Robinson was authorized to continue his 
field work until June, 1914, thus giving him practically a year in 
the field. It was planned that he should also extend his field 
work to neighboring islands, and at the time of his death he had 
made arrangements to visit Buru Island, as for the season he 
had secured a high percentage of the Amboina species to be 
found in flower or fruit. It was fully realized that his time 
could be more profitably spent in exploring neighboring islands, 
utilizing the intermediate periods between trips for a reexamina- 
tion of the various parts of Amboina for the purpose of locating 
in flower or fruit those species that had not been detected during 
his first period, July to December, of intensive field work. His 
wholly unexpected death prevented the fulfillment of these plans. 

This work has been based on the material and observations 
secured in a period of four and one-half months. It is evident 
that could the revised plan have been carried out and field work 
extended until June, 1914, much more material and data would 
have been available for study, with the result that the interpreta- 
tion of the Herbarium Amboinense would have been more satis- 
factory and more nearly complete than it is. 

Botanists and collectors who have actually prosecuted field 
work in Amboina* include LaBillardiere, the first naturalist to 
visit the island after Rumphius's death, Christopher Smith, the 
younger Roxburgh, Lahaie, Reinwardt, d'Urville, Zippel, Lesson, 
Hombron, Forsten, de Vriese, Teysmann, Naumann, Binnendyk, 
de Fretes, Beccari, Forbes, Warburg, Karsten, Boerlage, Treub, 
J. J. Smith, and Robinson. Some were there but for a few days, 
others for longer periods; and their collections, now widely 
scattered in different herbaria, comprise several thousand speci- 
mens. Were it possible to segregate from the various herbaria 



* Warburg, O. Die botanische Erforschung der Molukken seit Rumpf's 
Zeiten. Rumphius Gedenkboek 1702-1902 (1902) 63-78. 



INTRODUCTION 



27 



all of the Amboina material extant, doubtless many other obscure 
points regarding Rumphius's species could be elucidated, which 
in the following critical consideration I have been obliged to 
interpret from published descriptions alone. Doctor Robinson's 
four and one-half months of field work in Amboina were insuf- 
ficient in which to secure the necessary material and data to settle 
all of the doubtful points in connection with the forms de- 
scribed by Rumphius from Amboina material alone, and he had 
no opportunity to visit neighboring islands to search for special 
material that might serve to determine the status of Rumphius's 
rather numerous extra-Amboina species. 

ERRORS IN THE INTERPRETATION OF RUMPHIAN SPECIES 

The early botanical authors, such as Linnaeus, Burman f ., Lour- 
eiro, Lamarck, and numerous others, had but a slight conception 
of the principles of geographic distribution of plants, and ac- 
cordingly in their reductions of Rumphius's species many grave 
errors were committed. Very often in the early literature one 
finds the illustrations of an Amboina plant quoted as an exact 
synonym of a species of Indo-China, when in reality the two are 
totally different and not infrequently have been found to repre- 
sent different genera. It is not at all certain that in quoting 
illustrations of various species as synonyms Linnaeus and his 
contemporaries and immediate successors intended them as exact 
synonyms ; it would seem, in many cases at least, that the cita- 
tions of illustrations as synonyms was intended to convey to other 
botanists some conception of what the species was like, and not 
necessarily to indicate that it was an exact equivalent of the 
species under which it was cited. 

In the first two or three decades following the death of 
Linnaeus systematists were conservative in the matter of describ- 
ing new species. There was a very strong tendency to refer 
specimens to species already named by Linnaeus, rather than 
to describe material, even from distant and relatively unknown 
parts of the world, as new. Thus we find Loureiro in his Flora 
Cochinchinensis, published in 1790, erroneously referring num- 
erous Cochin-China specimens to Linnean species and likewise 
attempting to match his Cochin-China material with the Amboina 
species described and figured by Rumphius, apparently on the 
assumption that if a plant grew in Cochin-China, it should also 
grow in Amboina. In Loureiro's work there are scores of cases 
where the Rumphian name and figure are quoted as an exact 
synonym of a Cochin-China species described by him as new. 



28 



RUMPHIUS'S HERBARIUM AMBOINENSE 



In very many cases Loureiro took his specific name from Rumph- 
ius, yet in not a single case is a species described by Loureiro 
to be interpreted by the reference to Rumphius, as his descrip- 
tions were not based on data supplied by the Herbarium 
Amboinense, but on actual specimens from Cochin-China or 
southern China. 

We find the same condition in Burman f., Flora Indica (1768), 
where Burman's conception of the species proposed was not 
gained from the Rumphian synonym cited, often the only one 
given, so much as from actual specimens from Java or from 
some other part of the Indo-Malayan region; in few cases are 
Burman's species, as published in his Flora Indica, to be typified 
by the Rumphian reference cited. In the early volumes of 
Lamarck's Encyclopedic we find likewise numerous cases where 
species actually described from specimens originating in the 
Mascarene Islands, in the Philippines, and in other regions 
remote from Amboina are supplied with a Rumphian synonym, 
which usually has proved to be misplaced. Error after error 
has crept into systematic botany by interpretation of species by 
a Rumphian synonym, wrongly placed, rather than by consulta- 
tion of the actual type specimen. These errors, once published, 
have been perpetuated by other authors, sometimes because of 
failure to interpret types properly, sometimes because of lack 
of interest in problems of nomenclature, sometimes because of 
non-accessibility of type specimens for purposes of comparison, 
and for other reasons. By way of illustration I need cite only 
one or two extreme cases. 

The type of Fagara triphylla Lam. is a Philippine specimen 
collected by Sonnerat, and a recent examination of it in Lamarck's 
herbarium at the Museum d'histoire Naturelle, Paris, shows 
it to be identical with the endemic Philippine Melicope luzonensis 
Engl. De Candolle, however, apparently interpreting Fagara tri- 
phylla Lam. chiefly from the Rumphian synonym, Ampacus 
angustifolius Rumph., cited by Lamarck in the original descrip- 
tion, transferred it to Evodia as Evodia triphylla DC. ; and later 
authors, also interpreting it from the Rumphian synonym, have 
given Evodia triphylla (Lam.) DC. a range extending from 
India to Japan southward through Malaya to New Guinea. In 
clearing up this question of synonymy * I have shown that 
Fagara triphylla ~LsLm.=Evodia triphylla ~DC.=Melicope triphylla 
Merr. is a species confined to the Philippines; that Evodia tri- 

* On the identity of Evodia triphylla DC. Philip. Journ. Sci. 7 (1912) 
Bot. 373-378. 



INTRODUCTION 



29 



phylla of various authors includes at least three distinct species in 
two different genera; and now the occurrence of true Ampacus 
angustifolius Rumph. in the Amboina collection shows that this 
Rumphian species, while a true Evodia, represents still another 
distinct species. Evodia triphylla DC. as interpreted by various 
authors has included at least four distinct species in two different 
genera. 

Another case is that presented by Ricinus mappa Linn., based 
wholly on Folium mappae Rumph. This is the basis of Maca- 
ranga mappa Muell.-Arg., Mueller extending the range of the 
species to the Philippines by the erroneous reduction of Croton 
grandifolius Blanco as a synonym. I have shown that Maca- 
ranga grandifolia (Blanco) Merr. is a species entirely distinct 
from Macaranga mappa Muell.-Arg., yet Pax and K. Hoffman* 
in their recent monograph of this group interpret Macaranga 
mappa (Linn.) Muell.-Arg. wholly from Philippine specimens, 
erroneously citing Croton grandifolius Blanco, Macaranga por- 
teana Andre, and Macaranga grandifolia Merr. as synonyms, 
and even figuring the species from Philippine material. A 
casual comparison of Philippine material with Rumphius's figure, 
the type of Ricinus mappa Linn.= Macaranga mappa Muell.-Arg., 
shows that two totally different species are involved. The occur- 
rence of typical Folium mappae Rumph. in Robinson's Amboina 
collections shows conclusively that I was correct in separating 
the Philippine form, that Mueller was wrong in reducing Ricinus 
grandifolius Blanco to Macaranga mappa, and that Pax and K. 
Hoffman were entirely wrong in their interpretation of Maca- 
ranga mappa Muell.-Arg. The two species involved are so 
entirely different that they belong in distinct sections of the 
genus. 

THE INTERPRETATION OF RUMPHIAN SPECIES AS TYPES 

In the interpretation of the species of older authors under 
which Rumphian names are cited as synonyms one point must 
constantly be kept in mind. This is, as to whether the species 
was based on an actual specimen in the hands of the author or, 
by citation, wholly on the Rumphian description and figure. In 
nearly every case it is possible to determine this point merely by 
an examination of the description, for even when no specimen is 
actually cited, if the species was based on an actual specimen, 
data are usually given that could not have been derived from 

* Euphorbiaceae-Acalypheae-Mercurialinae. Engl. Pflanzenreich 63 (1914) 
320. 



30 ^ RUMPHIUS'S HERBARIUM AMBOINENSE 

either Rumphius's description or figure. Even in Linnaeus's 
works descriptions based on actual specimens rather than on 
cited synonyms are usually thus determinable. I have already 
noted that none of Loureiro's species, even when the specific 
name is taken from Rumphius, are to be interpreted by the 
Rumphian synonym cited. The same is true of most of 
Burman's species published in his Flora Indica, many of those 
proposed by Lamarck, and those of many other authors. 
Where a species was based on an actual specimen supplemented 
by a reference to Rumphius, the specimen is manifestly 
the type, but it then becomes necessary to determine ( whether 
or not the specimen represents the same species as the Rumphian 
synonym cited. In a very high percentage of such cases the 
actual specimen described has been found to represent a species 
different from the one figured by Rumphius, due to the fact 
that the early authors, having little conception of the geographic 
distribution of plants, failed to distinguish between the in- 
digenous and endemic elements in the Amboina flora and those 
species of wide distribution. Among all of the earlier workers 
there was a strong tendency to refer the Rumphian illustrations 
to species described from actual specimens, even if there was 
only a superficial resemblance between the specimen and the 
figure. None of them realized the necessity of interpreting 
Moluccan species from Moluccan specimens ; and, even if the value 
of such procedure were realized, no botanical material from 
Ambonia was available to European botanists until the close of 
the eighteenth century and, even then, only a limited amount. 

In the present consideration of the Herbarium Amboinense 
those species and their synonyms that were based solely on plants 
described and figured, or merely described, by Rumphius have 
been indicated by the term "type!" in parentheses following the 
citation. The list of such species could doubtless have been 
extended if in the course of the preparation of the manuscript, 
I had had access to all the literature. As it is, nearly 350 such 
4 'types" have been indicated. From the standpoint of taxonomy 
then, the Herbarium Amboinense is of relatively very great im- 
portance, for its descriptions and figures typify a very large 
number of binomials of later authors. Only two other pre- 
Linnean works on the Indo-Malayan flora can be compared with 
the Herbarium Amboinense in this respect, these being Rheede's 
Hortus Malabaricus and Linnaeus's Flora Zeylanica, and most of 
the actual specimens on which the later work was based are 
extant. 



INTRODUCTION 



31 



THE INTERPRETATION OF THE SPECIES DESCRIBED IN THE HERBARIUM 
AMBOINENSE BY VARIOUS AUTHORS 

LINNAEUS AND STICKMAN 

Citations of Rumphian synonyms are found in the first pub- 
lished work on the binomial system,* but these are few and 
of slight importance, as Linnaeus did not secure a copy of the 
Herbarium Amboinense until his manuscript was completed: 

Upus eximium beati Rumphii cura amplissimi D. Burmanni orbi botanico 
redditum, ad me accessit primum absoluto a typographo opere, cujus itaque 
synonymia seorsim tradere animus est.f 

About nineteen Rumphian synonyms are cited in the first 
edition of the Species Plantarum, of which four are in volume 
one, the remainder in volume two, but in only four cases are the 
Linnean species based wholly on the Rumphian synonyms, and 
these are all found in the Appendix. The list is as follows : 

Cynometra cauliflora Linn. Areca catechu Linn. 
Cynometra ramiflora Linn. , Caryota urens Linn. 

Averrhoa carambola Linn. Plukenetia volubilis Linn. 

Averrhoa bilimbi Linn. Hibiscus surattensis Linn. 

Garcinia mangostana Linn. Rubus parvifolius Linn. 

Acrostichum siliquosum Linn. Rubus moluccanus Linn. 

Borassus flabellifer Linn. Convolvulus peltatus Linn. 

Corypha umbraculifera Linn. Quercus molucca Linn. 

Cycas circinalis Linn. Croton variegatus Linn. 
Cocos nucifera Linn. 

The last four are based wholly on Rumphius's figures and 
descriptions. The new genus and species, Rumphia amboinensis 
Linn., is not based on Amboina material, contains no reference 
to the Herbarium Amboinense, and is the Indian Cordia tiliae folia 
(Poir.) Warb.J 

Linnaeus realized the great importance of Rumphius's work 
and at once assigned to one of his students, Olaf Stickman, a 
study of the Herbarium Amboinense. In the following year, 
1754, Stickman's dissertation on the Herbarium Amboinense was 
printed, this being probably the first work published following 
the binomial system of nomenclature after the system was pro- 
posed. Stickman's publication is discussed below. 

In the second edition of the Species Plantarum § Rumphius's 
Herbarium Amboinense is listed among the "Auctores Reforma- 

* Linnaeus, C. Species Plantarum (1753) 1-1200. 

f Linnaeus, C. Op. cit. 1199. 

JRumphius Gedenkboek 1702-1902 (1902) 78. 

§ Linnaeus, C. Species Plantarum ed. 2 (1762-63) 1-1684. 



32 



RUMPHIUS'S HERBARIUM AMBOINENSE 



tores" and is placed in the group of "Fundadores." Numerous 
reductions of Rumphian species had already been made in Stick- 
man's dissertation on the Herbarium Amboinense (1754), in the 
1759 reprint of this work,* and in the tenth edition of the Systema 
Naturae (1759) , most of which are repeated in the second edition 
of the Species Plantarum; in all there are about 275 references 
to the Herbarium Amboinense in this work. Additional refer- 
ences are included in Linnaeus's later works. 

The first attempt to interpret the species described and figured 
in the Herbarium Amboinense as a whole in terms of the binomial 
system of nomenclature was made one year after the system was 
proposed. This was nominally the work of Olaf Stickman, one 
of Linnaeus's pupils, but it is manifest that the actual work was 
largely that of Linnaeus himself. Stickman's dissertation^ as 
originally published, is a very rare work, and copies of it are 
known in but few libraries. Rouffaer and Muller J cite only 
two copies, one in the University Library at Upsala, Sweden, 
and one in the Konigliche Hof- und Staatsbibliothek, Munich. 
There is a copy in the library of the Linnean Society, London; 
one in the library of the British Museum; one in the library of 
the New York Botanical Garden; and one in the private library 
of Dr. J. H. Barnhart, of the New York Botanical Garden ; and I 
was so fortunate as to secure a copy from a European dealer for 
our work on the Amboina project. Doubtless other copies will 
be found in the older European libraries. 

In this work an attempt was made to reduce only those species 
figured by Rumphius in volumes one to six of the Herbarium 
Amboinense; volume seven, the Auctuarium, was not printed 
until the year following the appearance of Stickman's work and 
is, accordingly, not included. The species are arranged in the 
Rumphian sequence, giving the Rumphian name, the numbers 
of the plates, and the binomial equivalents of the various species 
so far as they could be determined. About three hundred of the 
forms figured by Rumphius are referred to binomials already 
proposed in the first edition of the Species Plantarum of Linnaeus, 

♦Amoen. Acad. 4 (1759) 112-143. 

t Herbarium Amboinense, | quod | consens. experient. Facult. Medicae 
| in Regia Academia Upsalensi, | sub praesidio | viri nobilissimi atque 
experientissimi, | Dn. Doct. Caroli | Linnaei, | . . . . | publico examini 
submittit, | Alumnus Regius I Olavus Stickman, | Smolandus. | In auditorio 
Car. Majori d X. Maji, | anno MDCCLIV | Upsaliae | . . . . (1754) 
IV + 1-28. 

t Eerste proeve van eene Rumphius-Bibliographie. Rumphius Gedenk- 
boek 1702-1902 (1902) 196. 



INTRODUCTION 



33 



or published for the first time in this dissertation. In addition 
to these three hundred specific reductions, many others are 
referred to generic names under the Linnean system, about 
twenty to species characterized by Rheede in his Hortus Mala- 
baricus, and a few are connected with other pre-Linnean names. 
As is to be expected many of the proposed reductions have 
since been shown to be wrong. 

In this work more than twenty binomials appear for the first 
time, and these, typified by the Rumphian figures and descrip- 
tions, although validly published, were overlooked by the com- 
pilers of Index Kewensis and do not appear in that work or, if 
included, are credited to later publications. Among these are 
the following : 

Garcinia celebica Linn. i Lens phaseoloides Linn. 

Psidium cujavus Linn. Menispermum flavum Linn. 

Myrtus leucadendra Linn. Adenanthera falcata Linn. 

Momordica indica Linn. Hernandia ovigera Linn. 

Plumbago indica Linn. Convallaria fruticosa Linn. 

Tragia scandens Linn. Piper decumanum Linn. 

Erythrina variegata Linn. Bromelia comosa Linn. 

Rhizophora caseolaris Linn. Dolichos pruriens Linn. 

Rhizophora corniculata Linn. Momordica trifolia Linn. 

Ricinus mappa Linn. Pothos latifolius Linn. 

Ricinus tanarius Linn. Pancratium narbonense Linn. 

Under modern conditions there would be no question whatever 
as to the authority for these names, for unless otherwise stated in 
the text the authority would be the author of the Dissertation. 
In this case the work was done under Linnaeus's inspiration and 
direction, and the reductions of the Rumphian species must have 
been made largely, if not wholly, by him. i have accordingly 
quoted Linnaeus as the authority for all new combinations 
appearing in Stickman's dissertation. 

In 1759 Stickman's Dissertation was reprinted under the title 
"Herbarium Amboinense, sub praesidio D. D. Car. Linnaei, pro- 
posuit Olavus Stickman, Smolandus"* Whatever doubt there 
may be as to the actual authorship of the original edition of 
Stickman's work in 1754, the 1759 reprint must certainly be 
credited to Linnaeus. This differs from the original edition 
notably in that the contents of volume seven of the Herbarium 
Amboinense, the Auctuarium, are included, while appended to the 
treatment of the Rumphian species is the "Flora Amboinensis," 
in which the species recognized are arranged under the Linnean 
classes Monandria, Diandria, etc., to which in turn an "Appendix" 



♦Linnaeus, C. Amoen. Acad. 4 (1759) 112-143. 

144971 3 



34 



RUMPHIUS'S HERBARIUM AMBOINENSE 



is added to include the Palmae and "Singulares." In the treat- 
ment of the species figured in volumes one to six of the Herbarium 
Amboinense, the 1759 reprint differs from the original edition 
in certain respects. Some corrections are made in the indicated 
binomials, and a few new ones are added. The number of 
Rumphian species reduced in this work is about 330, an increase 
of about 30 over the original edition, but this includes 12 from 
the Auctuarium that were not included in the 1754 edition. As 
in the original edition, certain binomials which have not been 
listed in Index Kewensis, appear in the 1759 reprint. Among 
these are Muntingia bartramia, Phaseolus cylindricus, Panicum 
vulpinum, Justicia bivalvis, Sesuvium portulacastrum, Varneria 
augusta, Lagerstroemia chinensis, and Canarium indicum. In 
the present work Linnaeus has been quoted as the authority for 
these binomials, as well as for those appearing in the original, 
1754, edition. 

BURMAN 

Burman, in editing the Herbarium Amboinense, added various 
notes on the identity of the Rumphian species and at the end of 
volume seven, the Auctuarium (1755), added his "Index univer- 
salis in sex tomos et auctuarium herbarii Amboinensis CI. Georgii 
Everhardi Rumphii," in which he reduced about 311 of the 
Rumphian species to the Linnean binomial system, for the most 
part following the reduction proposed by Linnaeus in Stickman's 
dissertation issued the preceding year. Here he also published 
a few new binomials typified by Rumphius's figures, which have 
been entirely overlooked by all later authors. Those in the 
vegetable kingdom are Mespilus silvestris Burm., Phaseolus 
marinus Burm., Pepo indicus Burm., and Aurantium maximum 
Burm. This index consists of twenty pages, unnumbered, the 
species being alphabetically arranged under their Rumphian 
names with references to the book, the chapter, and the volume 
in which they are described. 

In 1769 a second edition of this index was issued by Burman, 
under the title — 

Index alter in omnes tomos Herbarii Amboinensis CI. G. Everhardi 
Rumphii quern de novo recensuit, auxit et emendavit Joannes Burmannus, 
M. D. Botanices Professor, Academiae Caesareae Naturae Curiosum, nec 
non Regiae Scientiarum Academiae Upsaliensis membrum. 

It consists of twenty-two unnumbered pages, folio, and is 
apparently a rare work. Rouffaer and Muller * in their biblio- 
graphy cite an example of it which they examined in the library 



* Rumphius Gedenkboek 1702-1902 (1902) 200. 



INTRODUCTION 



35 



of the K. Svenskt Vetenskapsakademien, Stockholm, and one in 
the private library of Doctor Greshoff, late director of the 
Colonial Museum at Haarlem, Holland ; they consider it remark- 
able that the work is lacking in such libraries as those of Leiden, 
Upsala, Halle, and Munich. There are copies in the libraries of 
the British Museum ; the Linnean Society, London ; and the Royal 
Botanic Garden, Kew, England ; and I have been supplied with a 
photostat copy of it by Dr. Walter T. Swingle, of the United 
States Department of Agriculture, Washington, D. C. The 
second edition differs from the original chiefly in the additional 
reductions included. A total of about 458 binomials appear in 
this work, nearly 150 more than in the first edition. The reduc- 
tions included are chiefly those indicated in the second and the 
third editions of Linnaeus's Species Plantarum and in the 
younger Burman's Flora Indica, the latter having been published 
in 1768. Two new binomials appear, Achyranthes spiciflora 
Burm. and Verbesina aquatica Burm., the former, from the 
reference given, manifestly a misprint for Acalypha spiciflora 
Burm. f. ; while of the four published in the first edition in 1755 
Aurantium maximum Burm. is eliminated in favor of Citrus 
decumana Linn. 

BUCHANAN-HAMILTON 

Doctor Francis Buchanan-Hamilton commenced a critical con- 
sideration of the Rumphian species, which, however, was never 
completed or, at least, except for the first two parts, never 
published. The first part is entitled A Commentary on the 
Herbarium Amboinense, Liber Primus; and the second part, 
A Commentary on the Second Book of the Herbarium Am- 
boinense.* The work is of considerable interest and value. 
Each species described by Rumphius is discussed to a greater 
or less length with critical notes on the identity of the individual 
species and with reasons for and against the various proposed 
reductions. Hamilton was handicapped by lack of knowledge 
of the Malayan flora and attempted to interpret the Rumphian 
species largely from his experience with the Indian flora. The 
various forms actually described from Indian material are usually 
specifically distinct from the Rumphian species under which they 
were placed; but Hamilton usually does not claim that the 
Rumphian species is identical with the one he describes; he 
merely says that the two resemble each other or are manifestly 
allied. In a few cases, in the first part of his work, he proposed 
binomials typified by the Rumphian species, but none of these 



*Mem. Wern. Soc. Edinburgh 5 (1826) 307-383; 6 (1832) 286-333. 



36 



RUMPHIUS'S HERBARIUM AMBOINENSE 



appear in the second part. Unlike his predecessors who had 
attempted to interpret the Rumphian species, he did not confine 
his comments to the species figured, but also attempted to account 
for those described but not figured. 

HENSCHEL 

Henschel's * attempt to interpret the species described by 
Rumphius is merely an enumeration of the Rumphian species by 
volumes in the order in which they appear in the Herbarium 
Amboinense, with their binomial equivalents, so far as deter- 
mined, in parallel columns. He considers chiefly the species 
figured by Rumphius. The enumeration, a mere compilation, is 
very faulty and untrustworthy, as is to be expected, for 
Henschel was not a botanist and had no personal knowledge of 
the Indo-Malayan flora. New binomials, with one or two excep- 
tions, do not occur in this work, nor is there any critical discus- 
sion of the various species. The Clavis Herbarii Amboinensis 
of Henschel in arrangement compares closely with the sequence 
of Rumphian species published in the present work at the end 
of the systematic enumeration (p. 511 to 547), except that 
through the work of numerous botanists on the Malayan flora 
a very much higher percentage of the Rumphian species can now 
be placed than Henschel found possible. Henschel's work is, 
in general, of very slight value, so far as it applies to the species 
Rumphius described and their binomial equivalents. 

HASSKARL 

The most pretentious attempt heretofore made to interpret 
the species described in the Herbarium Amboinense is that by 
Doctor J. K. Hasskarl, entitled Neue Schliissel zu Rumph's 
Herbarium amboinense.f In this work the species are arranged 
under the Rumphian names in Rumphius's sequence, citing the 
page and plate numbers for each, the native names, and a 
chronological arrangement of the reductions proposed by various 
authors, with citations of literature. Hasskarl's work is scarcely 
more than a compilation, but is a remarkable example of patience 
and perseverance on the part of the author. The mere matter of 

*Henschel, A. G. E. T. Clavis Herbarii Amboinensis pp. 139-202, in his. 
Vita G. E. Rumphii, Plinii indici accedunt specimen materiae Rumphianae 
medicae clavisque herbarii et thesauri Amboinensis. (1833) XIV + 1-216. 

t Abhandlungen der Naturf. Gesellschaft Halle 9 (1866) 145-389; Re- 
print (1866) 1-247. 



INTRODUCTION 



37 



searching the literature from 1753 to 1866 for references to the 
Herbarium Amboinense must have entailed many months of 
exacting labor. Where numerous synonyms are cited, or at least 
numerous names are listed, to which a Rumphian species has 
been reduced, usually no opinion is expressed as to which is the 
correct one. Many of those suggested by Hasskarl himself are 
palpably wrong, due perhaps to his lack of knowledge of the 
Malayan flora. It is not evident that Hasskarl ever had a very 
wide knowledge of the flora of the Malay Archipelago in spite 
of his residence in Java and his published botanical contributions. 
Many of his errors of interpretation were primarily due to the 
same factor that caused others to fail in properly interpreting 
Rumphius's species; that is, a lack of botanical material from 
Amboina and the neighboring islands. On account of the method 
of arranging his data, Hasskarl's work is difficult to consult, 
is entirely unsatisfactory in aiding the botanist to gain a definite 
idea of which species are actually included in Rumphius's work 
and which are not, and because of the numerous errors in reduc- 
tions is, it is feared, more or less discredited among botanists 
familiar with the Malayan flora. 

Hasskarl's work performed one distinct service that his pre- 
decessors failed in. Stickman, Linnaeus, Burman, and Henschel 
dealt only or largely with the species figured by Rumphius, 
ignoring the descriptions to a large extent; Hasskarl, however, 
brought out clearly the fact that Rumphius described very many 
more forms than he figured. He perhaps went to extremes in 
enumerating all the variants of such plants as the coconut palm, 
sugar cane, rice, banana, and other cultivated forms, and cer- 
tainly went to extremes in attempting to reduce the Rumphian 
descriptions of these variants to named forms and varieties 
under the binomial system. Several binomials were proposed 
by Hasskarl, typified by citation of Rumphius's descriptions and 
figures, in attempting to account for Rumphius's species. These 
were overlooked in compiling Index Kewensis, but invariably 
fall as synonyms. 

In consulting Hasskarl's work, it should be noted that the 
numerous citations of Loureiro, Flora Cochinchinensis, are of 
the second, or Willdenow's, edition, 1793; that the references 
to Linnaeus are not to the original works of this author, but to 
Richter's Codex Botanicus Linneanus (1840) ; while Stickman's 
1754 dissertation on the Herbarium Amboinense, and usually also 
the 1759 reprint, is not cited. 



38 



RUMPHIUS'S HERBARIUM AMBOINENSE 



THE PRESENT STATUS OF RUMPHIAN SPECIES 

Rumphius named and described approximately 1,700 plants 
that he considered to represent distinct forms. However, many 
of the plants he named and characterized are "forms'' or 
"varieties" rather than "species" in the generally accepted sense 
? of these words. Slight variations in the color of the leaves, of 
the flowers, or of the stems of plants ; equally slight differences 
in the size of certain parts; and other trivial characters were 
deemed by him to be of sufficient importance to warrant the 
characterization of the form and the bestowal of a distinctive 
name. Thus, in the case of cultivated plants, such as the coconut, 
the betel-nut palm, the sago palm, the sugar cane, taro, rice, and 
balsam, both slight and prominent variants were distinguished, 
while in wild plants equivalent distinctions were often made. 

In terms of the binomial system, as species are understood 
to-day, the 1,700 forms named and to a greater or less degree 
characterized by Rumphius can be reduced to about 1,200 species, 
including those that, while apparently distinct, are of more 
or less doubtful status and have not been definitely referred to 
any genus. Of these 1,200 species about 930 can be definitely or 
fairly definitely referred to binomials, and about 140 additional 
ones can be safely placed in their respective genera, leaving 
about 130 that from data and material at present available 
cannot be definitely located under the binomial system ; some of 
these cannot be even placed in their proper families. A high 
percentage of these doubtful species are those that are very im- 
perfectly and briefly described, some being scarcely more than 
casually mentioned; few of them are figured. 

As already noted, many binomials have been based wholly on 
the Rumphian figures and descriptions. In about 800 cases 
references to the Herbarium Amboinense are found in the 
original descriptions or publications of species, while about 350 
binomials have been based wholly on various species more or less 
imperfectly characterized by Rumphius. During the past one 
hundred and thirty-five years numerous botanists have attempted 
with greater or less success to interpret these Rumphian species 
by connecting the Rumphian names and descriptions with actual 
botanical specimens. Many errors in interpretation and in iden- 
tification have been made, but the general results have been such 
that to-day a high percentage of the Rumphian species have been 
definitely connected with extant botanical material, and their true 
status has been determined. As a result of Doctor Robinson's 



INTRODUCTION 



39 



work in Amboina, the list of doubtful species has been greatly 
reduced. 

In the present consideration about 930 of the Rumphian species 
have been definitely referred to binomials, and of these about 
470 are definitely represented by botanical specimens collected 
by Doctor Robinson. It should be borne in mind, however, that 
many of the species that are not represented by specimens 
collected by Doctor Robinson were originally described by 
Rumphius from material that did not originate in Amboina, 
much of it coming from distant lands (see p. 14) . 

There are about 45 species, proposed by various authors, that 
are known to-day only from the data originally given by 
Rumphius; that is, those species that have not been, to my 
knowledge, definitely and correctly connected with actual bota- 
nical specimens from which in turn their true characters and 
relationships can be determined. Among these species of 
doubtful status are the following: 



Lentinus tuber regium Fries. 
Lentinus djamor Fries. 
Agaricus moschocaryanus Strinz. 
Polygaster sampadarius Fries. 
Pachyma tuber regium Fries. 
Pachyma hoelen Fries. 
Pandanus repens Miq. 
Pandanus baggea Miq. 
Pandanus terrestris Warb. 
Freycinetia graminea Blume. 
Bambusa excelsa Miq. 
Bambusa (Schizostachyum) longi- 

nodis Miq. 
Livistona bissula Mart. 
Calamus graminosus Blume. 
Calamus rumphii Blume. 
Calamus pisicarpus Blume. 
Calamus buroensis Mart. 
Calamus equestris Willd. 
Calaw v us cawa Blume. 
Calamus acidus Becc. 
Daemonorops niger Blume. 
Alpinia uviformis Horan. 
Michelia tsiampacca Linn. 
Polyalthia sp. (Uvaria ligularis 

Lam.). 



Polyalthia sp. (Guatteria rumphii 
Blume) . 

Goniothalamus sp. (Uvaria tripetala 

Lam.) . 
Talauma rumphii Blume. 
Mangifera utana Ham. 
Mangifera tapia Ham. 
Dillenia serrata Thunb. 
Sindora galedupa Prain. 
Actinodaphne rumphii Blume. 
Actinodaphne moluccana Blume. 
Canarium balsamiferum Willd. 
Canarium hirsutum Willd. 
Canarium zephyrinium Blume. 
Canarium sp. (Canariopsis pauci- 

juga Miq.). 
Canarium sp. (Pimela caryophyllacea 

Blume) . 
Osmoxylon umbelliferum Merr. 
Schefflera sp. (Brassaia littorea 

Seem.). 
Panax anisum DC. 
Hoya alba Kostel. 
Hoya elegans Kostel. 
Ipomoea rumphii Miq. 



Doubtless some of these species are represented in herbaria 
rich in Malayan material, but so far as published records go, 
such specimens have never been definitely connected with the 



40 



RUMPHIUS'S HERBARIUM AMBOINENSE 



published species that they may represent. The importance of 
interpreting species based on Rumphius's descriptions from 
actual specimens collected as near the classical locality as possible 
cannot be overestimated, and the sooner the above doubtful 
species are definitely connected with botanical material by which 
their true characters can be determined, the nearer we will be 
to the long hoped for stable nomenclature. 

DIFFICULTIES IN THE PROPER INTERPRETATION OF RUMP H IAN 

SPECIES 

The difficulties involved in attempting to interpret the species 
described by such an author as Rumphius in terms of the 
binomial system are very great. The actual working up of the 
Amboina collection has involved two entirely different sets of 
identifications, first an identication with the form Rumphius 
described or described and figured ; and, second, a further identi- 
fication of the same specimen to its proper genus and species 
under the binomial system. Neither task has been an easy one, 
for very obvious reasons. 

In dealing with the Rumphian descriptions, many difficulties 
are encountered. While often very long, the descriptions are 
nontechnical, and measurements are largely approximate or com- 
parative. TJie parts of the flowers are not described in detail, 
and often they are not even mentioned. The plants described 
in a single chapter under a "generic" term may belong to a single 
genus, as the term is understood to-day, or may belong in entirely 
different genera in distinct or even unrelated families. Many 
forms are only casually described, sometimes scarcely more 
than mentioned, while of others the description is reduced to 
a general description of the wood only. Very many of these 
casually described species were not based on Amboina material, 
but on specimens transmitted to Rumphius from various parts 
of Asia and Malaya. To a certain degree we have succeeded 
definitely in placing a high percentage of the species that are 
amply described and figured and a fair percentage of those that 
are but casually mentioned, but much remains to be done on 
this subject. 

Another factor that has rendered identifications difficult or 
uncertain is the figures themselves. While many of them are 
excellent and can be unmistakably referred to their proper species 
in the binomial system from an examination of the figures alone, 
others are very crude ; some are imperfect in that they delineate 
only leaf specimens ; some are manifestly based on material origi- 
nating from entirely distinct species or even from representatives 



INTRODUCTION 



41 



of different genera and families, and some do not agree at all 
with the descriptions to which they are ascribed. As already 
noted the artist has frequently depicted the leaves on one scale 
and the attached inflorescences, flowers, or fruits, as the case 
may be, on an entirely different scale. Very frequently the 
leaves are reduced in size, while the other parts may be greatly 
enlarged. In consulting the Herbarium Amboinense, it should 
be borne in mind that Rumphius himself never saw the figures, 
which were drawn by various artists after he became blind (see 
p. 16). 

Rumphius's idea of the species was not at all that of the species 
as understood to-day, nor can his chapter heads be considered 
as corresponding to the modern conception of the genus. As 
noted by Doctor Robinson in one of his letters to me : 

Rumphius imbibed the native ideas on the relationships of plants, and 
did his best to improve on them. Now the natives here to-day, and I 
think certainly also in his time, base their opinions largely on habit and 
leaf characters, or perhaps on habitat; thus mangi-mangi covers the whole 
mangrove family (Rhizophoraceae) with Sonncratia thrown in. Also to 
the characters utilized by the natives in making identifications should be 
added wood characters, latex if any, taste and smell of leaves, flowers, and 
fruit. Neither he nor they appreciate the primary value of flowers or fruit 
or of compound leaves. Again the methods of distinguishing species that 
we use were entirely unknown to him, as they are to the natives here to-day. 
We are so accustomed to putting emphasis on simple versus compound and 
opposite versus alternate leaves; superior versus inferior ovary; and apet- 
alous, polypetalous, and gamopetalous flowers and the number of their 
parts, that it is difficult to follow a man who took no count of any of these 
characters, except as to the compound leaves, while his opposite leaves are 
often opposite leaflets. He says in one place that a menispermaceous 
plant "maxime convenit" with what proves to be Derris uliginosa of the 
Leguminosae; what then about some of the other plants he described that 
"maxime convenit," when there is no illustration to suggest the identity 
of the species involved? Take the case of Ternstroemia, Ichthyoctonos 
montana of Rumphius. It is most excellently described and the illustration 
is fair, yet in this chapter he describes three forms which differ in the color 
of the wood and of the roots. It is incredible that in an island of this 
size that there can be three species of this small and characteristic genus 
to each of which the description can correspond so far as it goes and yet 
be worthy of being interpreted as three distinct species of Ternstroemia. 
There are two possible conclusions regarding it, and many other similar 
cases, first, that there is really only one species of Ternstroemia and that the 
differences are merely superficial; and, second, that he had in mind three 
really different species, not unlikely in as many different families of plants, 
but that the detailed description applies to one only; the other two forms 
briefly mentioned in this chapter are inextricable with certainty. Even if 
a sufficiently perfect knowledge of all the plants found in Amboina did 
enable us correctly to guess what was intended by the second and third 
forms of Ichthyoctonos, there is nothing in Rumphius's statements by 



42 



RUMPHIUS'S HERBARIUM AMBOINENSE 



which the correctness of the interpretation could be checked. Again Afa- 

cuerus is divided into "mas" and femina;" one is a Cyrtandra, of the 
Gesneriaceae, and the other is a Pellionia, of the Urticaceae, but he almost 
certainly included in the latter an equally common Elatostema. Conoceph- 
alus, of the Moraceae, and Medinilla, of the Melastoraataceae, are placed 
together. It will take much critical work certainly to distinguish in the 
Herbarium Amboinense such dissimilar plants as Pipturus, of the Urti- 
caceae, Zizyphus, of the Rhamnaceae, other melastomataceous plants in- 
cluding some species of Medinilla, Celtis, of the Ulmaceae, and even 
Strychnos, of the Loganiaceae. 

Very many similar cases could be added, but the above state- 
ment clearly indicates one particular phase of the difficulties 
involved in the interpretation of Rumphian species. 

The difficulties involved in identifying material under the 
binomial system have been very real. The herbarium of the 
Bureau of Science contains only such material as could be 
accumulated by actual field work and by exchanges in the past 
fifteen years, and while it contains a very fine series of Philippine 
species and much valuable material from the Indo-Malayan 
region generally, many species that I should like to have seen are 
lacking. Identifications, other than of those species already 
familiar to me, have been largely made by comparisons with the 
published descriptions, and very many such descriptions are 
entirely inadequate, especially those of the early authors. When- 
ever possible the original descriptions have been examined, but 
a number of works that it has been desirable or essential to 
examine in the course of the preparation of this manuscript are 
not available in Manila. In very numerous cases resource has 
been had to transcriptions or photographic reproductions of 
essential descriptions, and such data have been supplied by 
various botanists in Europe and America. In one form or 
another I have thus been able to examine nearly all of the 
references to Rumphius cited in this work. 

In the present consideration of the Rumphian species I have 
departed radically from the works of previous authors. In 
order to make the work more generally available to botanists, 
the Rumphian species, so far as they can be reduced at present, 
are cited as synonyms under the various species and genera to 
which they refer, these again being arranged by families and 
genera in the sequence of Engler and Prantl's Natiirlichen 
Pflanzenfamilien. Appended to this systematic treatment of 
the Rumphian species is a list under the Rumphian names 
arranged in the sequence of the Herbarium Amboinense, giving 
references to the volume, the page, and the figure under each 
and, so far as determinable, their binomial equivalents. 



INTRODUCTION 



43 



NOMENCLATURE 

In nomenclature the rules of the Vienna Botanical Congress * 
have been closely followed, including the list of nomina conser- 
vanda as well as the supplementary list adopted by the Brussels 
Botanical Congress. f The sole exception in the list of nomina 
conservanda is the adoption of the generic name Taetsia in place 
of Cordyline for what are considered to be entirely valid reasons. 

Up to the close of the last century comparatively little attention 
was given to the question of priority in the names of plants, 
and many authors accepted or changed generic and specific 
names at will. It is true that in a majority of cases names 
well established were generally accepted, but changes were often 
made for the most trivial reasons. In work prosecuted under 
these lax but easy methods of selecting names for plants, the 
exact identity of obscure species was a matter of relatively 
slight importance. 

With the establishment and general acceptance of the prin- 
ciple of priority in selecting the names of species, it has become 
important, from the viewpoint of stability of nomenclature, to 
determine so far as possible the exact status of the species 
proposed by the older authors. It would admittedly be con- 
venient if many of the names proposed by early authors could 
be discarded, but if we ignore the species of one author, any 
botanist at any time would be justified in likewise ignoring 
species proposed by any other author, which would result in 
a veritable chaotic condition in nomenclature. We can no 
longer look on the work of this or that author, no matter how 
incomplete or imperfect, as unworthy of consideration, nor can 
we accept Hooker's J dictum, regarding species proposed by 
such authors as Blanco, that it was undesirable to devote time 
to their identification. 

Regarding species based on such pre-Linnean works as 
Rheede's Hortus Malabaricus and Rumphius's Herbarium Am- 
boinense, Hooker f. commends the work of Blume for the good 
service he has performed to the antiquarian branch of botany 
in interpreting Rumphian species. The general adoption of the 
principle of priority has emphasized the great importance of 
what Hooker f. characterized as the antiquarian branch of 

* Briquet, J. Regies internationales de la Nomenclature botanique 
adoptees par le Congres International de Botanique de Vienne 1905 (1906) 
1-99. 

t Actes du III nie Congres International de Botanique, Bruxelles, 1910 1 
(1910) 112-116. 

$ Hooker f. & Thomson. Flora Indica 1 (1855) Introductory Essay 56. 



44 



RUMPHIUS'S HERBARIUM AMBOINENSE 



botany, and while investigations of the status of binomials 
proposed by the early authors will lead to necessary changes 
in nomenclature, and changes in the accepted names of plants 
are always to be regretted, yet a strict interpretation of species 
from a historical standpoint will correct numerous current 
misapplications of names and lead to the assignment of these 
names to the forms for which their original authors intended 
them. 

The strict application of the rule of priority as to the specific 
names has resulted in many changes in nomenclature, but 
these changes are unavoidable, if the international code be 
followed. Considering the distinctly basic position occupied 
by the Herbarium Amboinense in Malayan botany, the fact that 
so many binomials based wholly on Rumphius's work have been 
published by various authors, and the further fact that a high 
percentage of the ' 'species" so established have been unintelligible 
to most botanists and have hence frequently been redescribed 
under other names, it is rather surprising that more changes in 
nomenclature have not been found necessary. The conservative 
botanist will be shocked to learn that as a result of the present 
investigation of the Herbarium Amboinense such common, widely 
distributed, and well-known species as the pineapple, the soy bean, 
the cow pea, and the pomelo must receive new specific names; 
that such names as Vigna luteola Sw., Canavalia turgida Grah., 
Pongamia glabra Vent., and others equally well known for the 
last hundred years or more fall as synonyms ; and that in the true 
mangrove trees (Rhizophoraceae) the specific names in current 
use for most species are wrongly applied. 

In proposing changes in nomenclature, I have not hesitated 
even when such well-known species as Glycine hispida, Ananas 
sativus, Citrus decumana, and Phaseolus unguiculatus are in- 
volved. While objections may be voiced to the changes in 
nomenclature herein proposed, while individual botanists may 
refuse to adopt the proposed changes, and while exceptions may 
be taken to some of my interpretations, I cannot understand why 
logically the changes proposed should not be generally adopted. 
Each case has been critically worked out from a historical 
standpoint, and the accepted name is that indicated by the 
International Code of Botanical Nomenclature now generally 
accepted by most botanists. To those botanists who make their 
own rules and in the matter of accepting or rejecting specific 
names are a law unto themselves, no appeal is made, for appeal 
is useless. I am firmly of the opinion, however, that stability in 



INTRODUCTION 



45 



nomenclature can come only by adhering to definite rules and 
by critically working out the proper name for each species in 
conformity with those rules. 

In one matter affecting generic nomenclature I have definitely 
gone on record in a previous publication.* This is that the 
generic name should be maintained for the group for which it 
was intended by its original author, not applied to representatives 
of a group that was wholly unknown to the author of the generic 
name. Nauclea of Linnaeus serves as an illustration of the idea. 
As Nauclea is currently interpreted, it contains nothing that was 
originally placed there by Linnaeus. I have proposed to apply 
Nauclea in its original sense; that is, to the species currently 
placed in Sarcocephalus and have proposed the new name 
Neonauclea for Nauclea auctt., not of Linnaeus. It is to me 
inconceivable that a genus proposed by one author should be 
interpreted by others with every original species excluded. The 
application of this principle to some of the older genera will 
involve an adjustment in such a case as Alpinia, for the type and 
sole species cited in the original description of Alpinia is a 
Renealmia. 

As a natural consequence of the acceptance of the International 
Rules of Botanical Nomenclature, the numerous Rumphian names 
adopted by 0. Kuntze f as substitutes for more ' 'recent" generic 
names of other authors have been wholly ignored. It is perfectly 
evident that Rumphius had no idea of the "genus," and his names 
cannot be interpreted in a generic sense. Even in the Auctuar- 
ium, which was published after the binomial system was 
established, the "generic" names certainly cannot be considered 
as the equivalent of the genus as understood by contemporary 
botanists. The binomials that appear in the Auctuarium (1755) 
are merely accidental and cannot be considered as properly 
"published" binomial names; Burman in his added notes oc- 
casionally cites the first edition of Linnaeus's Species Plantarum, 
but he never cites the Linnean binomial, merely the descriptive 
sentence. It is perfectly clear that he had no intention of pub- 
lishing the Rumphian accidental binomials as binomials in the 
Linnean sense. It is only in his Index Universalis that he rec- 
ognized the Linnean system. In this he reduced to binomials 
those Rumphian species whose status had been determined by 
Linnaeus and by Stickman and proposed a few new binomials. 

* On the application of the generic name Nauclea of Linnaeus. Journ. 
Wash. Acad. Sci. 5 (1915) 530-642. 

fRev. Gen. PI. (1891-93) CLV + 1-1011. 



46 



RUMPHIUS'S HERBARIUM AMBOINENSE 



OVERLOOKED BINOMIALS 

During the course of my work on this project I have detected 
an unusually large number of binomials, over one hundred, 
that while validly published were overlooked in compiling 
the data for Index Kewensis and do not appear in that 
work or in any of its supplements. Some of these binomials 
appear in works that were never indexed, such as Stickman's 
Dissertation on the Herbarium Amboinense (1754) ; and in 
various other works based on the Herbarium Amboinense, such 
as those of Henschel and of Hasskarl; while others appear in 
works that were indexed, but for one reason or another certain 
species were overlooked. The list has been deemed to be of 
sufficient importance to warrant publication and is accordingly 
here presented. 

Achyranthes chinensis Osbeck Dagbok Ostind. Resa (1757) 205 = ? A. 
aspera Linn. 

Achyranthes spicifiora Burm. Index Alt. Herb. Amb. (1769) [b]=Acalypha 

amentacea Roxb. 

Afzelia rhomboidea F.-Vill. Novis. App. Fl. Filip. (1880) 72=Pahudia 
rhomboidea Prain. 

Allaeanthus luzonicus F.-Vill. Novis. App. Fl. Filip. (1880) 198. No com- 
bination in "Benth. et Hook. f. Gen. III. 361." 

Allophylus grossedentatus F.-Vill. Novis. App. Fl. Filip. (1880) 51 (Schmi- 
d.elia grossedentata Turcz.). 

Alpinia rufa Naves Novis. App. Fl. Filip. (1880) 226 (Hellenia rufa Presl). 

Alpinia scabra Naves Novis. App. Fl. Filip. (1880) 226 (Hellenia scabra 
Blume) . 

Amomum hatuanum Naves Novis. App. Fl. Filip. (1880) 224=? Amomum 
aculeatum Roxb. 

Anona sariffa Roxb. ex Henschel Vita Rumph. (1833) 142=Diospyros kaki 
Linn. f. 

Artocarpus fretisii Teysm. & Binn. ex Hassk. in Abh. Naturf. Gesellsch. 

Halle 9 (1866) 189. 
Aurantium maximum Burm. Index Univers. Herb. Amb. Auct. (1755) 

il6]=Citrus maxima (Burm.) Merr. (C. decumana Linn.). 
Bombax aculeatum Linn. Syst. ed. 10 (1759) 1141 = Ceiba pentandra 

Gaertn. 

Breweria alsinoides F.-Vill. Novis. App. Fl. Filip. (1880) 143=Jacque- 

montia paniculata Hallier f. 
Briza elegans Osbeck Dagbok Ostind. Resa (1757) 246 = ? Eragrostis 

elegantula Steud. 

Bromelia comosa Linn, in Stickman Herb. Amb. (1754) 21; Amoen. 
Acad. 4 (1759) 130= Ananas comosus (Linn.) Merr. (A. sativus 
Schult. f.). 

Bulbophyllum carinatum Naves Novis. App. Fl. Filip. (1880) 235 (Epiden- 

drum carinatum Linn.). 
Bulbophyllum purpureum Naves Novis. App. Fl. Filip. (1880) 234 (Sarco- 

podium purpureum Reichb. f.). 



INTRODUCTION 



47 



Bursera ? nitida F.-Vill. Novis. App. Fl. Filip. (1880) 41 (Marignia nitida 

Turcz. ) = Glycosmis cochinchinensis Pierre. 
Cadamba nocturna Ham. ex Henschel Vita Rumph. (1833) Ihl—Nauclea 

undulata Roxb. 

Caesalpinia jayabo Maza in Anal. Soc. Esp. Hist. Nat. 19 (1890) 234. 
Canarium indicum Linn. Amoen. Acad. 4 (1759) 134 p. ^. — Canarium 

commune Linn. 
Canarium zephyrinum Blume Mus. Bot. 1 (1850) 217. 

Caryophyllus silvestris Teysm. ex Hassk. in Abh. Naturf. Gesellsch. Halle 

9 (1866) Vol— Eugenia caryophyllata Thunb. 
Caryota javanica Osbeck Dagbok Ostind. Resa (1757) 210 — Ceratolobus 

javanicus (Osbeck) Merr. (C. glaucescens Blume). 
Cassumbium spinosum Ham. ex Henschel Vita Rumph. (1833) 143=Schlei- 

chera oleosa (Lour.) Merr. 
Catesbaea ? javanica Osbeck Dagbok Ostind. Resa (1757) 92 — Cleroden- 

dron commersonii Spreng. 
Citrus grandis Osbeck Dagbok Ostind. Resa (1757) 9S = Citrus maxima 

(Burm.) Merr. (C. decumana Linn.). 
Citrus limonia Osbeck Reise Ostind. China (1765) 250 (C. limonium 

Risso) . 

Citrus sinensis Osbeck Dagbok Ostind. Resa (1757) 250. 

Codiaeum bractiferum Roxb. Fl. Ind. ed. 2, 3 (1832) 680. 

Columnea chinensis Osbeck Dagbok Ostind. Resa (1757) 230 = Limnophila 

chinensis (Osbeck) Merr. (L. hirsuta Benth.). 
Commelina chinensis Osbeck Dagbok Ostind. Resa (1757) 2A2 — Commelina 

nudiflora Linn. 

Convallaria chinensis Osbeck Dagbok Ostind. Resa (1757) 219=&c^a 
chinensis Benth. 

Convolvulus indicus Burm. Index Univers. Herb. Amb. Auct. (1755) [6] = 

Ipomoea indica (Burm.) Merr. (/. congesta R. Br.). 
Cordia tiliaefolia Warb. in Rumphius Gedenkboek (1902) 78 (Rumphia 

amboinensis Linn., R. tiliaefolia Poir.). 
Costus ananassae Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 333 = 

Tapeinochilus ananassae K. Schum. 
Cryptanthus chinensis Osbeck Dagbok Ostind. Resa (1757) 277 (quid?). 
Cucumis rumph ii Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 280 = 

Cucumis sativus Linn. 
Curculigo rumphiana Schultes Syst. 7 (1830) lbl = Curculigo orchoides 

Gaertn. 

Cycas pectinata Ham. in Mem. Wern. Soc. 5 (1826) 322. 

Dehaasia borneensis F.-Vill. Novis. App. Fl. Filip. (1880) 179 (Haasia 

borneensis Meisn.). 
Desmodium cumingianum F.-Vill. Novis. App. Fl. Filip. (1880) 61 (Den- 

drolobium cumingianum Benth. 
Dioscorea nummularifolia Henschel Vita Rump. (1833) 183 sphalm=.D. 

nummularia Lam. 
Donax canniformis K. Schum. in Engl. Bot. Jahrb. 15 (1893) 440. 
Eleocharis dulcis Trin. ex Henschel Vita Rumph. (1833) 186. 
Erndlia subpersonata Giseke Prael. Ord. Nat. PI. (1792) 252 = Curcuma 

longa Linn. 

Eroteum lanigerum Blanco Fl. Filip. (1837) H0 = Trichospermum lani- 
gerum (Blanco) Merr. (T. trivalve Merr.). 



48 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Eugenia longiflora F.-Vill. Novis. App. Fl. Filip. (1880) 86 (Syzygium 

longiflorum Presl). 
Fimbristylis cumingii F.-Vill. Novis. App. Fl. Filip. (1882) 308 =Bulbostylis 

barbata Kunth. 

Flindersia radulifera Spreng. Gesch. Bot. 2 (1818) 76 ex Hassk. in Abh. 

Naturf. Gesellsch. Halle 9 (1866) 206=FUndersia amboinensis Poir. 
Govantesia malulucban Llanos in Rev. Progr. Cienc. 15 (1865) 191 = 

Champereia manillana (Blume) Merr. 
Guatteria rumphii Blume ex Henschel Vita Rumph. (1833) VbZ—Polyal- 

thia sp. 

Habenaria cordata Naves Novis. App. Fl. Filip. (1880) 250, non R. 

Br., = Habenaria diphylla Dalz. 
Hapaloceras ? arupa Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 

193 — Payena leerii Kurz. 
Hibiscus convoivulaceus Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 

7 4 = Hibiscus surattensis Linn. 
Hibiscus haenkeanus F.-Vill. Novis. App. Fl. Filip. (1880) 25= A belmoschus 

haenkeanus Presl. 

Homalium aranga Vidal ex F.-Vill. Novis. App. Fl. Filip. (1880) 94, in 

syn. = Homalium luzonicum F.-Vill. 
Hypoestes cumingiana F.-Vill. Novis. App. Fl. Filip. (1880) 157. No 

combination in "Benth. et Hook. f. Gen. II. 122." 
Ichnocarpus acuminatus F.-Vill. Novis. App. Fl. Filip. (1880) 131 = 

Aganosma acuminata G. Don. 
Ichnocarpus macrocarpus F.-Vill. Novis. App. Fl. Filip. (1880) 131= A ga- 

nosma macrocarpa A. DC. 
Ichnocarpus velutinus F.-Vill. Novis. App. Fl. Filip. (1880) 131= Aganosma 

velutina A. DC. 

Lagerstroemia chinensis Linn. Amoen. Acad. 4 (1759) 137= Lag erstroemia 
indica Linn. 

Lagurus paniculatus Linn, ex Burm. f. Fl. Ind. (1768) 30 — Andropogon 
nardus Linn. 

Legnotis lanceolata Blanco Fl. Filip. (1837) 445 =Decaspermum panicu- 
latum Kurz. 

Macanea arborea Blanco Fl. Filip. (1837) 431 = Alphonsea arborea 

(Blanco) Merr. (A. philippinensis Merr.). 
Macrolobium amboinense Teysm. ex Hassk in Abh. Naturf. Gesellsch. 

Halle 9 (1866) 189 =Intsia bijuga O. Kuntze. 
Mangifera utana Ham. in Mem. Wern. Soc. 5 (1826) 326. 
Medinilla lagunae Vidal ex F.-Vill. Novis. App. Fl. Filip. (1880) 89, descr. 
Melaleuca trinervis Ham. ex Henschel Vita Rumph. (1833) 145 = Mela- 
leuca leucadendra Linn. 
Melia parasitica Osbeck Dagbok Ostind. Resa (1757) 277=? Lansium 

domesticum Correa. 
Mespilus sylvestris Burm. Index Univers. Herb. Amb. Auct. (1755) [14] = ? 

Carissa carandas Linn. 
Mespilus sylvestris Burm. Index Univers. Herb. Amb. Auct. (1755) [18] = 

Flacourtia indica (Burm.) Merr. 
Milium zonatum Llanos Frag. PI. Filip. (1851) 24:=Eriochloa ramosa O. 

Kuntze. 

Melochia indica A. Gray ex F.-Vill. Novis. App. Fl. Filip. (1880) 29 = 

Melochia umbellata Stapf. 



INTRODUCTION 



49 



Mimosa chinensis Osbeck Dagbok Ostind. Resa (1757) 2SS=Albizzia 

chinensis (Osbeck) Merr. (^4. stipulata Boiv.). 
Momordica indica Linn, in Stickman Herb. Amb. (1754) 24, Amoen. 

Acad. 4 (1759) 132 = Momordica charantia Linn. 
Monarda chinensis Osbeck Dagbok Ostind. Resa (1757) 240 (quid?). 
Moringa domestica Ham. in Mem. Wern. Soc. 5 (1826) 268, Sll=Moringa 

oleifera Lam. 

Muntingia bartramia Linn. Amoen. Acad. 4 (1759) 124 = Commersonia 

bartramia (Linn.) Merr. (C. platyphylla Anders.). 
Murraya scandens Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 

233 = M. paniculata Jack. 
Nauclea elegans Teysm. & Binn. ex Hassk. in Abh. Naturf. Gesellsch. 

Halle 9 (1866) 100 =Anthocephalus macrophyllus Havil. 
Negretia pruriens Blanco Fl. Filip. ed. 2 (1845) 441 =Mucuna pruriens DC. 
Oberonia ancipita Naves Novis. App. Fl. Filip. (1880) 230 (sphalm! O. 

anceps Lindl.). 

Octomeles moluccana Teysm. & Binn. ex Hassk. in Abh. Naturf. Gesellsch. 

Halle 9 (1866) 208 = Octomeles sumatrana Miq. 
Orchis lanigera Blanco Fl. Filip. (1837) 611=Rhynchosia retusa Blume. 
Osmelia philippica F.-Vill. Novis. App. Fl. Filip. (1880) 93 (Stachycrater 

philippinus Turcz., Osmelia philippinensis Benth.). 
Panax rumphii Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 78 = 

Nothopanax tricochleatum Miq. 
Pancratium narbonense Linn, in Stickman Herb. Amb. (1754) 28=Eury- 

cles amboinensis Lindl. 
Panicum philippinum F.-Vill. Novis. App. Fl. Filip. (1882) S12=Axonopus 

semialatus Hook, f . 
Panicum tuberosum Llanos Frag. PI. Filip. (1851) 4:l = Panicum repens 

Linn. 

Panicum vulpinum Linn. Amoen. Acad. 4 (1759) 134=Setaria flava Kunth. 
Pepo indicus Burm. Index Univers. Herb. Amb. Auct. 7 (1755) [6]=Cwcwr- 
bita pepo Linn. 

Phaseolus cylindricus Linn. Amoen. Acad. 4 (1759) 132= Vigna cylindrica 
(Linn.) Merr. 

Pholidocarpus rumphii Meisn. ex Hassk. in Abh. Naturf. Gesellsch. Halle 

9 (1866) 12 = Pholidocarpus ihur Blume. 
Phytolacca ? javanica Osbeck Dagbok Ostind. Resa (1757) 21Q = Termi- 

nalia catappa Linn. 
Pimelandra disticha F.-Vill. Novis. App. Fl. Filip. (1880) 123= Ardisia 

disticha A. DC. 

Platanthera horsfieldii Naves Novis. App. Fl. Filip. (1880) 251 (Peristylus 
gracilis Blume). 

Plumbago indica Linn, in Stickman Herb. Amb. (1754) 24, Amoen. Acad. 

4 (1759) 133 (P. rosea Linn., 1762). 
Polyscias cumingiana F.-Vill. Novis. App. Fl. Filp. (1880) 102 (Paratropia 

cumingiana Presl) . 
Pterocaulon redolens F.-Vill. Novis. App. (1880) 116. 
Pycnanthemum decurrens Blanco Fl. Filip. ed. 2 (1845) 333=Hyptis 

capitata Jacq. 

Randia racemosa F.-Vill. 'Novis. App. Fl. Filip. (1880) 108 (Stylocoryne 
racemosa Cav.). 

Raton la montana F. Vill. Novis. App. Fl. Filip. (1880) 52=Arytera 
montana Blume. 

144971 4 



50 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Raton ia rufescens F.-Vill. Novis. App. Fl. Filip. (1880) 52 (Zygolepis 

rufescens Turcz.) =Arytera litoralis Blume. 
Rhizophora longissima Blanco Fl. Filip. (1837) 27S=R. mucronata Lam. 
Ricinus ruber Miq. Fl. Ind. Bat 1 2 (1858) 390 —Ricinus communis Linn. 
Samyda trivalvis Blanco Fl. Filip. (1837) 37 4 = Casearia. 
Sarcochilus centipeda Naves Novis. App. Fl. Filip. (1880) 238 (Thrixsper- 

mum centipeda Lour.). 
Sarcochilus tenuifolius Naves Novis. App. Fl. Filip. (1880) 238 (Epiden- 

drum tenuifolium Linn.). 
Saurauia elegans F.-Vill. Novis. App. Fl. Filip. (1880) 19 (Scapha elegans 

Choisy, Saurauia rugosa Turcz.). 
Solidago chinensis Osbeck Dagbok Ostind. Resa (1757) 2Al = Wedelia 

calendulacea Less. 

Spermacoce discolor F.-Vill. Novis. App. FL Filip. (1880) 113 (Borreria 
discolor DC.) . 

Tapeinochilus ananassae K. Schum. in Engl. Bot. Jahrb. 27 (1899) 249. 
Tetradapa javanorum Osbeck Dagbok Ostind. Resa (1757) 93— Erythrina 
indica Lam. 

Timonius nitidus F.-Vill. Novis. App. Fl. Filip. (1880) 109 (Petesia nitida 
BartL). 

Timonius ternifolius F.-Vill. Novis. App. Fl. Filip. (1880) 109 {Petesia 
ternifolia Bartl.). 

Torenia glabra Osbeck Dagbok Ostind. Resa (1757) 210 (Torenia bentham- 
iana Hance). 

Tragia scandens Linn, in Stickman Herb. Amb. (1754) 12, Amoen. Acad. 
4 (1759) 128 = Tetracera scandens (Linn.) Merr. (Tetracera sarmen- 
tosa Vahl. 

Tylophora bifida F.-Vill. Novis. App. Fl. Filip. (1880) 134 (Oxystelma 
bifidum Llanos) . 

Vallisneria sphaerocarpa Blanco Fl. Filip. (1837) 780= Enhalus acoroides 
Steud. 

Varneria augusta Linn. Amoen. 4 (1759) 136 = Gardenia augusta (Linn.) 

Merr. (G. florida Linn.). 
Verbesina aquatica Burm. Index Alt. Herb. Amb. (1769) [18']=Wedelia 

biflora DC. 

Zizyphus littorea Teysm. ex Hassk. in Abh. Naturf. Gesellsch. Halle 9 
(1866) 116—Ximenia americana Linn. 

ACKNOWLEDGMENTS 

We are under great obligations to the Dutch colonial officials 
for various courtesies extended to Doctor Robinson, but more 
especially are we indebted to Dr. J. C. Koningsberger, director 
of the botanic garden at Buitenzorg, Java, and to Dr. J. J. 
Smith of the same institution for their interest in the Amboina 
project and their hearty cooperation. Through their interest 
Doctor Robinson was supplied with a copy of the Herbarium 
Amboinense for his use during the period of his field work 
in Amboina, with a portion of his field equipment, and with 
the services of the mantri Mardjoeki, a Javanese assistant, 
who aided in the field work and had general oversight of the 



INTRODUCTION 



51 



drying of all material collected. I am also indebted to Doctor 
Koningsberger for the original map of Amboina, on which the 
one presented herewith was based, and for numerous specimens 
representing species not available in Manila, but which it was 
desirable to examine. 

Assistance has been had from several specialists in the pre- 
paration of this report. The treatment of the Marantaceae, 
of the Zingiberaceae, and of Heliconia, in the Musaceae, as 
presented in this work, is that of Doctor Th. Valeton; the 
Orchidaceae is the work of Doctor J. J. Smith. The other 
groups have been worked up by me, but I have had the assistance 
of various specialists in certain families. The Pteridophyta 
have been determined by Captain C. R. W. K. van Alderwereldt 
van Rosenburg, of Buitenzorg.* The Pandanaceae were deter- 
mined by Doctor U. Martelli, Florence, Italy; the Palmae by 
Doctor 0. Beccari, Florence, Italy; the Bambusae by J. Sykes 
Gamble, esq., East Liss, England; the Piperaceae by Mr. C. de 
Candolle, Geneva, Switzerland; and the Sapindaceae by Doctor 
L. Radlkofer, Munich, Germany. Doctor Th. Valeton has 
assisted me in the identification of the Rubiaceae, while Mr. 
F. S. Collins, North Eastham, Massachusetts, has identified the 
algae and has kindly supplied me with extracts from books 
not available in Manila regarding the Rumphian species of 
this group. Mr. I. H. Burkill, director of the Botanic Garden, 
Singapore, has supplied me with critical notes regarding Dios- 
coreaceae. To Doctor Walter T. Swingle, Washington, D. C, 
I am indebted for a photostat copy of Burman's "Index alter," 
a work not available in Manila and of which I was unable to 
secure a copy; to Mr. F. V. Coville, Mr. P. L. Richer, and Mr. 
S. C. Stuntz, of the United States Department of Agriculture, 
I am indebted for typewritten or photostat copies of numerous 
original descriptions not available in Manila; to Doctor George 
T. Moore, director of the Missouri Botanical Garden, St. Louis, 
Missouri, for the loan of certain books ; to Sir David Prain and 
Mr. A. W. Hill, of the Kew gardens, London, England, and to 
Doctor A. B. Rendle, of the British Museum, for copies of de- 
scriptions and for critical notes on various type specimens. To 
all of these gentlemen I wish to express my thanks for assistance 
rendered, without which the present consideration of the 
Rumphian species must of necessity have been less complete and 
more inexact than it is. 



* The Amboina Pteridophyta collected by C. B. Robinson. Philip. Journ. 
Sci. 11 (1916) Bot. 101-123, t. 5, 6. 



SYSTEMATIC ENUMERATION 



THALLOPHYTA 

Rumphius described a small number of thallophytes, and 
fortunately but few of the forms he described and figured 
have been made the types of species under the binomial system. 
The algae are represented by some of the more prominent forms, 
such as Sargassum, Turbinaria, and a few of the Rhodophyceae, 
but, in all, Rumphius characterized and named only about a dozen 
species of this group; none of these have been made the types 
of species under the binomial system. Of the fungi more num- 
erous forms were described and figured, about twenty-five being 
characterized. Some of these are wholly unrecognizable from 
the descriptions alone, while the identity of others is perfectly 
evident. About ten of the forms Rumphius described have been 
made the types of species under the binomial system, most of 
these binomials typified by Rumphius's figures and descriptions 
being proposed by Fries. In the lichens but two forms were 
described; both are apparently referable to Usnea. 

In the Bryophyta not a single species was described, unless 
some of the pendant epiphytic mosses and hepatics were in- 
cluded in the rather generalized description of Muscus capillaris 
Rumph. Herb. Amb. 6: 89, t UO, f. 2; the figure, and the descrip- 
tion at least for the most part, refers to Usnea rather than to 
any of the bryophytes. 

ALGAE * 

CHLOROPHYCEAE 

CHAETOMORPHA Kiitzing 

CHAETOMORPHA JAVANICA Kiitz. 

Capillus nympharum Rumph. Herb. Amb. 6: 90, 179. 

Rumphius's description applies to Chaetomorpha, but the form 
he had might be either C. javanica Kiitz. or the allied C. 
brachygona Harv. (Ayer putri, Rel. Robins. 2393, locally known 
as lumu-lumu) , which Doctor Robinson states is "not unlikely 
Capillus nympharum:' Martens states that it is an "Algen aus 

* In the following consideration of the algae described by Rumphius, the 
reductions in general follow G. von Martens in Die Preussische Expedition 
nach Ost-Asien. Botanische Theil. Die Tange (1866-68) 1-152, t. 1-8. I 
am indebted to Mr. F. SI Collins, of North Eastham, Massachusetts, for 
copies of passages in Martens's work pertaining to the Herbarium Amboin- 
ense and for identifying the Amboina algae collected by Doctor Robinson. 

53 



54 



RUMPHIUS'S HERBARIUM AMBOINENSE 



der Familie Conf erven und, der Localitat nach zu schleissen, 
zunacht Chaetomorpha javanica Kiitz., welche mein Sohn an 
der angegebene Stelle weider gesamelt hat." 

RHODOPHYCEAE 

GELIDIUM Lamouroux 

GELIDIUM AM ANSI I Kiitz. 

Muscus gelatinus japponensis Rumph. Herb. Amb. 6: 90. 

In the text, page 88, t. UO, f. 3, is referred to Muscus gelatinus 
japponensis, and this figure is placed with the description by 
Hasskarl, Neue Schlussel (1866) 167; to me the figure looks 
distinctly like Gracilaria lichenoides (Linn.) Harv. Martens 
refers the figure to Chaetomorpha javanica Kiitz., where it 
certainly does not belong. Henschel referred the description 
and the figure to Sphaerococcus gelatinus Ag., while Hasskarl 
suggested that both might be referable to Eucheuma spinosum 
Ag. Martens states that the species, as described, "ist sicher 
das von Siebold aus Japan mitgebracht Gelidium Amansii Kiitz., 
aber Rumph. scheint verschiedene andere heterogene Gegen- 
stande nicht gehorig davon zu unterscheiden." 

HIMANTHALIA Lyngbye 

HIMANTHALIA LOREA (Linn.) Lyngb. 

Bodelha altera Rumph. Herb. Amb. 6: 187. 

This is scarcely more than mentioned by Rumphius as growing 
on the coasts of Spain and Portugal. The reduction follows 
Martens. 

GRACILARIA Greville 

GRACILARIA LICHENOIDES (Linn.) Harv. 

Sphaerococcus lichenoides Ag. 

Alga coralloides I Rumph. Herb. Amb. 6: 181, t. 7k, /. 8, t. 76, f. A-C. 

Amboina, Hitoe lama, Robinson PL Rumph. Amb. 573, November 5, 1913, 
on coral in shallow water. 

Loureiro, Fl. Cochinch. (1790) 686, referred this to Lichen 
rocella Lour., which is an entirely erroneous disposition of it, 
whatever Loureiro's species may prove to be. It has been 
referred by various authors to Fucus edulis Gmel. and to Sphae- 
rococcus lichenoides Agardh, the former a synonym of the latter. 
Martens considers it to be Sphaerococcus lichenoides Agardh= 
Gracilaria lichenoides Harv. The illustration given on t. U0 f 
f. 3, referred in the text to Muscus gelatinus japponensis, and by 
Martens referred to Capillus nympharum Rumph. ==C7&ae£o- 
morpha javanica Kiitz., very much better agrees with named 



PHAEOPHYCEAE 



55 



specimens of Gracilaria lichenoides (Linn.) Harv., than do the 
figures cited above, t 74, f. 3, t. 76, f. A-C. 

PHAEOPHYCEAE 

SARGASSUM Agardh 

SARGASSUM POLYCYSTUM J. Ag. 

Acetabulum marinum Rumph. Herb. Amb. 6: 185, t. 76, f. 1. 

This was referred by Burman f., Fl. Ind. (1768) 239, to Fucus 
natans Linn., where it manifestly does not belong; Henschel 
referred it to Sargassum amboinicum "Rumph." ; and Hasskarl, 
Neue Schliissel (1866) 181, placed it under Sargassum myrio- 
cystum J. Ag. The present reduction follows Martens. 

SARGASSUM BACCIFERUM Ag\ 

Sargassum pelagicum Rumph. Herb. Amb. 6: 188, t, 76 f. 2. 

Linnaeus originally reduced this to Fucus natans Linn., in 
Stickman Herb. Amb. (1754) 28, Amoen. Acad. 4 (1759) 136, 
Syst. ed. 10 (1759) 1345; Burman f., Fl. Ind. (1768) 239, placed 
it under Fucus granulatus Linn.; and Henschel placed it under 
Sargassum bacciferum Agardh, where it properly belongs. The 
only forms of Sargassum collected by Doctor Robinson in 
Amboina are referable to S. binderi Sonder. 

SARGASSUM AQUIFOLIUM (Turn.) J. Ag. 

Agarum III funiculare s. foliatum Rumph. Herb. Amb. 6: 186. 

In the very short description given by Rumphius three distinct 
forms are mentioned and casually described. According to 
Martens the first one is Carpacanthus herbaceus K\itz.=Sargas- 
sum aqui folium J. Ag. He suggests that the other forms may 
be referable to Sargassum granuliferum Agardh. 

SARGASSUM FLAVIFOLIUM Kiitz. 

Sargasso s. Wier Rumph. Herb. Amb. 6: 167. 

This is briefly mentioned as growing along the coasts of Spain 
and Portugal. The reduction follows Martens. 

TURBINARIA Lamouroux 

TURBINARIA ORNATA (Turn.) J. Ag. 

Acetabulum marinum inf undibuliforme Rumph. Herb. Amb. 6: 185. 
Amboina, Paso, Robinson, PI. Rumph. Amb. 576, October 29, 1913, 
washed ashore, locally known as arien. 

Henschel placed this under Sargassum turbinatum Agardh, 
but Hasskarl, Neue Schliissel (1866) 181, and Martens placed 



56 RUMPHIUS'S HERBARIUM AMBOINENSE 

it under Turbinaria ornata (Turn.) J. Ag., quoting J. Agardh 
Sp. Alg. 1 : 266. This is manifestly the correct disposition of it. 

TURBINARIA sp.? 

Acetabulum marinum e Macassar Rumph. Herb. Amb. 6: 186. 

Hasskarl, Neue Schlussel (1866) 181, thought that this might 
be referable to Turbinaria vulgaris J. Ag., var conoidea J. Ag., 
but Martens expresses the opinion that it might perhaps be the 
same as Chauvinia macrophysa Kutz.=Caulerpa racemosa var. 
clavifera forma macrophysa (Kiitz.) Weber. The description is 
very short and indefinite. 

FUCUS Linnaeus 
FUCUS VESICULOSUS Linn. 

Bodelha Rumph. Herb. Amb. 6: 187. 

The common rock weed is briefly mentioned by Rumphius as 
growing on the coasts of Spain and Portugal. 

MASTOCARPUS Kiitzing 

MASTOCARPUS KLENZEANUS Kiitz. 

Agarum II s. bracteatum Rumph. Herb. Amb. 6: 186. 

Henschel referred this to Fucus bracteatus Ag., where it cer- 
tainly does not belong. Martens considers that it is certainly 
referable to Mastocarpus klenzeanus Kiitz. 

ALGAE OF ENTIRELY DOUBTFUL STATUS 

Alga coralloides sinensium Rumph. Herb. Amb. 6: 90. 

This is scarcely more than casually mentioned, and the identity of the 
plant intended is entirely problematical, other than that it is a marine alga. 

Agarum lactucarium Rumph. Herb. Amb. 6: 186. 

The brief description includes several entirely different forms of Rhodo- 
phyceae, none of which are certainly determinable. Martens suggests that 
the last mentioned may be Hypnea divaricata J. Ag. 

Agarum corticosum Rumph. Herb. Amb. 6: 187. 

This is wholly indeterminable from data at present available, and no 
author has as yet suggested a possible reduction of it. The description is 
very brief and imperfect. 

FUNGI 
LENTINUS Fries 
LENTINUS SAJOR CAJU Fries Epicr. (1836-38) 393 (type!). 
Agaricus sajor caju Fries Syst. 1 (1821) 175 (type!). 
Boletus primus infundibuli forma [figura] Rumph. Herb. Amb. 
6: 125, t. 56, f. 1. 

Amboina, Gelala, Robinson PI. Rumph. Amb. 571, September 19, 1913, 
on old tree trunks, altitude about 175 meters. 



FUNGI 



57 



The Rumphian figure and description are the whole basis of 
Lentinus sajor caju Fries, the species apparently having been 
generally interpreted correctly, as the figure is quite char- 
acteristic. Philippine material referred by Bresadola to Len- 
tinus sajor caju differs but slightly from the Amboina specimen 
cited above. The form briefly described in the second para- 
graph, indicated by Hasskarl as forma altera varietas, while 
undoubtedly a Lentinus, may or may not be the same as Lentinus 
sajor caju Fries. 

LENTINUS TUBER REGIUM Fries Epicr. (1836-38) 392 (type!); Sac- 

cardo Syll. Fung. 1 (1887) 604. 
Agaricus tuber regium Fries Syst. 1 (1821) 174 (type!). 
Tuber regium Rumph, Herb. Amb. 6: 120, t. 57, f. U (cf. Pachyma 

tuber regium Fries p. 61). 

A species imperfectly understood, based wholly on Rumphius's 
figure and description, although perhaps correctly interpreted 
by Hennings.* The subterranean portion of Tuber regium, as 
figured and described by Rumphius, is Pachyma tuber regium 
Fries, a species of wholly doubtful status (see p. 61), and 
is apparently nothing but a pseudo-sclerotium of Lentinus tuber 
regium Fries. f 

LENTINUS DJAMOR Fries Epicr. (1836-38) 395 (type!). 
Agaricus djamor Fries Syst. 1 (1821) 185 (type!). 
Boletus II arboreus Rumph. Herb. Amb. 6: 125, t. 56, f. 2, 3. 

A species known only from Rumphius's description and rather 
crude figures, but probably correctly placed in the genus 
Lentinus. 

M ARASM I US Fries 

MARASMIUS sp. 

Cassutha cornea Rumph. Herb. Amb. 7: 52. 

Rumphius's description applies to a plant that can hardly 
be other than the mycelium of one of the horsehair blights, 
Marasmius sp., for a general consideration of which see Petch.J 
Burman f., Fl. Ind. (1768) 93, referred it to Cassytha corniculata 
Burm. f., a species described and figured from Javan specimens, 
and which is perhaps a species of Galeola of the Orchidaceae. 
Linnaeus, Mant. 2 (1771) 237, repeats Burman's description 
under Cassyta corniculata, the reduction being cited by Loureiro, 
Murray, Lamarck, and Willdenow. Miquel, Fl. Ind. Bat. I 1 

* Engl. & Prantl. Nat. Pflanzenfam. 1 1 : 225; see Lloyd, Myc. Notes 47 
(1917) 666, fig. 959. 

t See Petch, T. The Pseudo-sclerotia of Lentinus similis and Lentinus 
infundibuliformis. Ann. Bot. Gard. Peradeniya 6 (1915) 1-17, t. 1. 

% Horse-hair blights. Ann. Bot. Gard. Peradeniya 6 (1915) 43-68, t. 2-7. 



58 



RUMPHIUS'S HERBARIUM AMBOINENSE 



(1858) 977, correctly excluded Cassytha corniculata Burm. f. 
from the Lauraceae. Hasskarl, Neue Schliissel (1866) 191, 
states: "Mihi Rhizomorpha aut Mycelium fungi cujusdam esse 
videtur." 

GANODERMA Karsten 

GANODERMA AMBOINENSE (Lam.) Pat. in Bull. Soc. Bot. Fr. 5 
(1889) 70. 

Agaricus amboinensis Lam. Encycl. 1 (1783) 51 (type!). 
Polypoms amboinensis Fries Syst. Mycol. 1 (1821) 354 (type!). 
Fomes amboinensis Fries Epicr. (1836-38) 442 (type!). 
Fungus elatus cochlearis Rumph. Herb. Amb. 6: 129, t. 57, f. 1. 
Amboina, Hitoe messen, Robinson PI. Rumph. Amb. 572, altitude about 
400 meters. 

The Rumphian figure and description are the whole basis of 
all the synonyms cited above. This is not Ganoderma amboin- 
ense (Lam.) Pat. as currently interpreted, but is apparently 
a form of Ganoderma rugosum (Bl. & Nees) Bres. In this 
connection it is to be noted that Rumphius figures Fungus elatus 
cochlearis with a long stipe and definitely states regarding it : 
"petiolo longo & tenui, spithamam vel pedam circiter longo," 
which includes no character of Ganoderma amboinense as cur- 
rently interpreted. The figure shows a specimen with a lateral 
pileus, while Ganoderma rugosum Bres. usually has a central 
stipe. Of Robinson's material, cited above, one specimen has 
a central stipe, and one, the pileus injured, has a lateral stipe. 
I have little hesitation in interpreting true Ganoderma amboin- 
ense (Lam.) Pat. as the form currently known as Ganoderma 
rugosum Bres. From this I do not think that Fungus elatus 
primus Rumph. and Fungus elatus petasoides Rumph., described in 
the preceding paragraph, can be distinguished. 

GANODERMA COCHLEAR (Nees) comb. nov. 

Polyporus cochlear Nees in Nov. Act. Acad. Nat. Cur. 13: 20, t. 6. 
Ganoderma amboinense auct. plur., non Agaricus amboinensis Lam., 

nec Polyporus vel Fomes amboinensis Fries. 
Fungus elatus digitatus Rumph. Herb. Amb. 6: 129, t. 57, /. 2, 3, et, 

s. n., t. 57, f. E. 

Amboina, Amahoesoe, Robinson PI. Rumph. Amb. 610, August 30, 1913, 
on dead trees at low altitudes. 

It is very evident from an examination of the original descrip- 
tions that Ganoderma amboinense Pat. has been wrongly inter- 
preted by recent authors — Patouillard, Murrill, Sydow, and 
Bresadola — for the Rumphian figure and description, on which 
Ganoderma amboinense is based, is undoubtedly the form 
currently known as Ganoderma rugosum Bres. I am of the 



FUNGI 



59 



opinion, however, that t. 57, f. 2, 3, represent juvenile forms 
of a Ganoderma, probably G. cochlear as here interpreted, and 
that figure E, described by Burman as "tarn cochlearis, quam 
digitati species est ex utrisque mixta," belongs with Ganoderma 
cochlear (Nees) Merr. ; that is, it is Ganoderma amboinense 
auct., non (Lam.) Pat. Lamarck considered figures 2 and 3 to 
represent a variety of his Agaricus amboinensis. Loureiro, Fl. 
Cochinch. (1790) 694, erroneously reduced these figures to 
Helvella mitra Linn., which is a totally different plant. It is not 
certain whether or not Polyp orus pisachapanni Nees is distinct 
from Ganoderma cochlear (Nees) Merr. 

FAVOLUS Fries 

FAVOLUS sp. 

Fungus arboreus III Rumph. Herb. Amb. 6: 128. 

The description is unmistakably referable to Favolus or some 
very closely allied genus, such as Hexagonia, as indicated by the 
brief description of the lower surface as "subtus autem in varias 
cellulas & tessaras distincta est instar favorum Apium." 

POLYSTICTUS Fries 

POLYSTICTUS SANGUINEUS (Linn.) Nees in Mey. Prim. Fl. Esseq. 
(1818) 304. 

Boletus sanguineus Linn. Sp. PI. ed. 2 (1763) 1646. 
Fungus arboreus II (ruber) Rumph. Herb. Amb. 6: 128. 

The phrase "utrimque rubra/' together with the few other 
characters given in the very short description, definitely refers 
the form Rumphius described to the strongly marked Polystictus 
sanguineus Nees. 

POLYPORUS Micheli 

POLYPORUS sp. 

Fungus arboreus II (albus) Rumph. Herb. Amb. 6: 128. 

The brief description applies to Polyporus or to one of the 
very closely allied genera. There is nothing sufficiently definite 
in the description to warrant even a guess at its identity. 

POLYPORUS sp. 

Fungus arboreus I Rumph. Herb. Amb. 6: 127. 

Indeterminable from the data given by Rumphius, other than 
that it is referable to Polyporus, sensu latiore, or to one of 
the very closely allied genera that have been segregated from it. 
Both Polyporus lucidus Fries and P. amboinensis Fries= Gano- 
derma amboinense Pat. have been suggested for the Rumphian 
species. 



60 RUMPHIUS'S HERBARIUM AMBOINENSE 

AGARICUS Linnaeus 

AGARICUS MOSCHOCARYANUS Streinz Nomencl. Fung. (1861) 70 
(type!). 

Boletus moschocaryanus Rosenthal Syn. PI. Diaphor. (1862) 31 
(type!). 

Boletus moschocaryanus Rumph. Herb. Amb. 6: 124. 

From Rumphius's description this can scarcely be an Agari- 
cus, but is more probably a Lentinus. Rumphius describes it 
as an edible fungus growing on Myristica trees. Its status as 
a species is quite unknown. 

HIRNEOLA Fries 

HIRNEOLA AURICULA J U DAE (Fries) Berk. Outl. (1860) 289. 
Tremella auricula Linn. Sp. PL (1753) 1157. 
Exidia auricula judae Fries Syst. 2 (1823) 221. 
Boletus V auris murina Rumph. Herb. Amb. 6: 126, t. 56, f. U> 

This was placed by Loureiro, Fl. Cochinch. (1790) 695, under 
Peziza auricula Lour. It can scarcely be other than the common 
and widely distributed Hirneola auricula judae Berk. 

AGARICACEAE sp. 

Boletus saguarius Rumph. Herb. Amb. 6: 124. 

The genus of this is uncertain, but it manifestly belongs in 
the Agaricaceae. Rumphius describes it as edible and as grow- 
ing on the decaying waste from the trunks of sago palms from 
which the sago has been extracted. 

AGARICACEAE sp. 

Boletus II umbraculiforma Rumph. Herb. Amb. 6: 126. 

All that can be said regarding this form is that it belongs in 
the Agaricaceae, possibly in the genus Agaricus. It is one of 
the edible forms. 

AGARICACEAE sp. 

Boletus IV terrestris Rumph. Herb. Amb. 6: 126. 

Of wholly doubtful status other than that it belongs in the 

Agaricaceae. 

AGARICACEAE sp. 

Fungus igneus Rumph. Herb. Amb. 6: 130, t. 56, /. 5. 

The description is not sufficiently definite to warrant even a 
guess as to the genus the plant pertains to, and the figure is very 
poor. It is nonedible and is stated by Rumphius to be poisonous. 

LYCOPERDON Tournefort 

LYCOPERDON sp. 

Crepitus lupi verus Rumph. Herb. Amb. 6: 131. 



FUNGI 



61 



Lycoperdon is the suggested reduction of this by Hasskarl, 
which may be the correct disposition of it. Fungus arborum tube- 
rosus Rumph., Herb. Amb. 6: 130, also may be possibly referable 
to the same genus. Beyond a surmise as to the genus, no further 
reduction can be suggested from data at present available. 

DICTYOPHORA Desvaux 

DICTYOPHORA PHALLOIDEA Desv. Journ. Bot. 2 (1809) 88. 

Hymenophallus indusiatus Vent, in Mem. Inst. Nat. Sci. 1 (1789) 
520. 

Phallus daemonum Fries Syst. Mycol. 2 (1823) 283 (type!). 
Hymenophallus daemonum Spreng. Syst. 4 (1827) 498 (type!). 
Dictyophora speciosa Klotzsch. in Nov. Act. Acad. Nat. Cur. 19 

(1843) Suppl. 1 : 239, t. 6. 
Phallus daemonum Rumph. Herb. Amb. 3: 218; 6: 131, t. 56, f. 7. 

The Rumphian figure is the type of Phallus daemonum Fries 
and of Hymenophallus daemonum Spreng., and it is certainly the 
same as Dictyophora phalloidea Desv. Loureiro, Fl. Cochinch. 
(1790) 694, erroneously reduced it to Phallus impudicus Linn. 
Ventenat's specific name is the oldest one, but no change is here 
proposed in the nomenclature of the species. 

POLYGASTER Fries 

POLYGASTER SAM PADARI US Fries Syst. Mycol. 2 (1823) 295 (type!). 
Tuber sampadarium Rumph. Herb. Amb. 6: 123. 

Loureiro, Fl. Cochinch. (1790) 697, referred Tuber sampa- 
darium Rumph. to Lycoperdon glomeratum Lour., a species based 
on Cochin-China specimens. There is no evidence that the form 
Loureiro described is the same as the one Rumphius considered. 
The Rumphian description typifies Polygaster sampadarius Fries, 
a species of very doubtful status. Fischer, in Engler & Prantl 
Nat. Pflanzenfam. I 1 ** (1899) 399, places Polygaster as a 
doubtful genus under the Plectobasidiineae (Sclerodermineae) . 

PACHYMA Fries 

PACHYMA TUBER REGIUM Fries Syst. Mycol. 2 (1823) 243 (type!). 
Tuber regium Rumph. Herb. Amb. 6: 120, t. 57, f. A. 

The genus Pachyma Fries is one of doubtful status, although 
there is little doubt that Pachyma tuber regium Fries is nothing 
but a pseudo-sclerotium of Lentinus tuber regium Fries (see 
Lentinus tuber regium Fries, p. 57). 

PACHYMA HOELEN Fries Syst. Mycol. 2 (1823) 243 (type!). 
Hoelen Rumph. Herb. Amb. 6: 122. 



62 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Rumphius's description of Hoelen was based on material 
originating in China. It is the whole basis of Pachyma hoelen 
Fries and, like Pachyma tuber regium Fries, is of doubtful status. 
It is cultivated on pine trees in various parts of China*" and 
has been referred to Pachyma cocos Fries. Specimens of fulin, 
kindly secured for me by Mr. W. J. Tutcher in a Chinese drug 
store in Hongkong, agree closely with the excellent figures of 
Pachyma cocos Fries given by Currey in Trans. Linn. Soc. 23 
(1860) t 10, f. 5, 6, 9. A part of Mr. Tutcher's specimen was 
sent to Dr. W. A. Murrill, of the New York Botanical Garden, 
who states that he has sclerotia of the same general type from 
different localities in America, but that the only method of 
distinguishing them accurately is to develop the fruiting form. 
In some cases the fruiting form proves to be species of Polyporus, 
in others species of Lentinus. He expresses the opinion that 
Pachyma hoelen Fries is distinct from P. cocos Fries. 

FUNGUS indet. 

Muscus frutescens III muscagineus Rumph. Herb. Amb. 6: 87. 

The description apparently applies to the mycelium of some 
fungus, but the status of Muscus frutescens muscagineus is 
wholly indeterminable. 

LICHENES 
USNEA Linnaeus 

USNEA sp. 

Barba saturni Rumph. Herb. Amb. 6: 88. 

Henschel thought that this might be a species of Lycopodium, 
which is an impossible reduction of it. Hasskarl, Neue Schlussel 
(1866) 167, states (( Usneae aut gen. aff. Lichenum spec, qua- 
edam." The form described is probably an Usnea. 

USNEA sp. 

Muscus capillaris Rumph. Herb. Amb. 6: 89, t. 40, f. 2. 

Linnaeus, in Stickman Herb. Amb. (1754) 27, Amoen. Acad. 
4 (1759) 135, Syst. ed. 10 (1759) 975, erroneously reduced this 
to Renealmia usneoides Linn., which is the American Tillandsia 
usneoides Linn., of the Bromeliaceae. Burman f., Fl. Ind. (1768) 
239, cites it under Lichen capillaris Burm. f ., of which, however, 
it is scarcely the type. Loureiro, Fl. Cochinch. (1790) 171, dis- 
cusses it under Grammica aphylla Lour. =Cuscuta chinensis Lam. 
and definitely refers it, op. cit. 687, to Lichen usnea Linn. 
Rumphius's description perhaps includes more than an Usnea, 



* Shaw, N. Chinese Forest Trees and Timber Supply (1914) 39, 295. 



CYATHEACEAE 



63 



possibly pendant epiphytic mosses and hepatics, but his figure, 
and his description at least in most part, apparently refers to 
Usnea or to some very closely allied genus. His material was 
from the higher mountains of the interior of Amboina. 

PTERIDOPHYTA 

The entire Amboina collection of this group, made by Doctor 
Robinson, has been critically studied by Capt. C. R. W. K. van 
Alderwerelt van Rosenburgh, of Buitenzorg, Java.* In pre- 
paring the present consideration of the species described by 
Rumphius I have had the benefit of his published work, both as to 
the names of the various species under the binomial system, and 
as to the identity of the forms Rumphius named and described. 
In a few cases I have made changes in nomenclature, but in no 
case, except Cyathea, involving new combinations. 

CYATHEACEAE 

CYATHEA Smith 

CYATHEA RUM PHI AN A (v. A. v. R.) comb. nov. 

Alsophila rumphiana v. A. v. R. in Philip. Journ. Sci. 11 (1916) Bot. 
104. 

Palmifilix alba Rumph. Herb. Amb. 6: 63. 
Amboina, Hitoe messen, Robinson PI. Rumph. Amb. 463, October 14, 
1913, in light forests, altitude about 175 meters. 

Doctor Robinson, who had the opportunity of examining this 
tree fern in the field and of making a direct comparison with 
Rumphius's description, considers this identification of Palmi- 
filix alba Rumph. as certain. 

CYATHEA AM BOI N ENSIS (v. A. v. R.) comb. nov. 

Alsophila amboinensis v. A. v. R. in Philip. Journ. Sci. 11 (1916) Bot. 
103. 

Palmifilix nigra Rumph. Herb. Amb. 6: 63 (t. 27?). 
Amboina, Hatiwe and'Soja, Robinson PI. Rumph. Amb. 464, 465, August 
and September, 1913, in forests, altitude 300 to 400 meters. 

The only previously suggested- reduction of Palmifilix nigra 
Rumph. is that of Henschel and Pritzel, who referred it to Cya- 
thea arbor ea Sm., a species of tropical America. The reduction 
made here is probably the correct disposition of Palmifilix nigra 
Rumph. The illustration, however, may belong to any one of 
the three forms described in this chapter, the particular one 
intended not being indicated by Rumphius. 

* The Amboina Pteridophyta collected by C. B. Robinson. Philip. Journ. 
Sci. 11 (1916) Bot. 101-123, t. 5, 6. 



64 



RUMPHIUS'S HERBARIUM AMBOINENSE 



CYATHEA sp. 

Palmifilix postium Rumph. Herb. Amb. 6: 63. 

Manifestly one of the tree ferns and probably a Cyathea. Its 
exact status cannot be determined from any data at present 
available. 

POLYPODIACEAE 

DRYOPTERIS Adanson 

DRYOPTERIS FEROX (Blume) 0. Kuntze Rev. Gen. PI. 2 (1891) 812. 
Aspidium ferox Blume Enum. PI. Jav. (1828) 153. 
Filix amboinica mas Rumph. Herb. Amb. 6: 69. 
Amboina, Robinson PL Rumph. Amb. 439, July 23, 1913, on river banks 
in the vicinity of the town of Amboina. 

This reduction of Filix amboinica mas is certainly correct, 
the species being a strongly marked and characteristic one and 
Rumphius's description agreeing closely with it. 

TECTARIA Cavanilles 
(Aspidium Swartz) 

TECTARIA CRENATA Cav. Descr. (1802) 250. 

Aspidium repandum Willd. Sp. PI. 5 (1810) 216. 

Lonchitis amboinica recta Rumph. Herb. Amb. 6: 70. 
Amboina, Soja, Robinson PL Rumph. Amb. 447, August 2, 1913, altitude 
about 250 meters. 

This is probably the correct disposition of the plant Rumphius 
briefly described. 

STENOSEM I A Presl 

STENOSEM I A AURITA (Sw.) Presl Tent. Pterid. (1836) 237. 

Acrostichum auritum Sw. in Schrad. Journ. 1800 2 (1801) 12. 
Filix florida Rumph. Herb. Amb. 6: 78, t. 35, f. 1. 

Amboina, Way uri, Robinson PL Rumph. Amb. 444, September 9, 1913, 
in river bottoms at low altitudes. 

This reduction was made by Willdenow, Sp. PI. 5 (1810) 112, 

as Acrostichum auritum Sw. It has been cited under Polybo- 
trya aurita Blume and Acrostichum floridum Poir., both of which 
are synonyms of Stenosemia aurita Presl. 

DAVALLIA Smith 

DAVALLI A ELATA (Forst.) Spreng. in Schrad. Journ. 17 9 9 2 (1799 ) 271. 

Trichomanes elatum Forst. Prodr. (1786) 85. 

Dryopteris triplex arborea Rumph. Herb. Amb. 6: 73, t. 32, f. 1. 
Amboina, Ayer putri, Robinson PL Rumph. Amb. 449, July 28, 1913, 
on trees at low altitudes. 



POLYPODIACEAE 



65 



Very closely allied to Davallia denticulata (Burm. f.) Mett., 
but considered by van Alderwereldt van Rosenburgh to be speci- 
fically distinct; see Philip. Journ. Sci. 11 (1916) Bot. 108. 

TAPEINIDIUM C. Christensen 

TAPEINIDIUM AMBOYNENSE (Hook.) C. Chr. Ind. Fil. (1906) 631. 
Davallia amboynensis Hook. Sp. Fil. 1 (1846) 178, t. 56. 
Dryopteris triplex silvestris I terrestris Rumph. Herb. Amb. 6: 73. 
Amboina, Lateri, Robinson PL Rumph. Amb. 443, September 9, 1913, 
in forests, altitude about 250 meters. 

This is probably the correct disposition of the Rumphian spe- 
cies. Blume thought it was a species of Aspidium, and Hasskarl 
placed it with doubt under Davallia patens Sw., to which it cer- 
tainly cannot be referred. 

ATHYRIUM Roth 

ATHYRIUM ESCULENTUM (Retz.) Copel. in Philip. Journ. Sci. 3 (1908) 
Bot. 295. 

Hemionitis esculenta Retz. Obs. 6 (1791) 38. 

Diplazium esculentum Sw. in Schrad. Journ. 1801 2 (1803) 312. 

Filix esculenta Rumph. Herb. Amb. 6: 67, t. 29. 

This characteristic, widely distributed, and well-known fern 
is not represented in our Amboina collections. The Rumphian 
figure is a good representation of the species and is unmistakably 
Athyrium esculentum Copel. Henschel and Pritzel have re- 
ferred it to Diplazium malabaricum Spreng., which is a synonym 
of Athyrium esculentum Copel. 

ASPLENIUM Linnaeus 
ASPLENIUM NIDUS Linn. Sp. PI. (1753) 1079. 

Phyllitis amboinica I arborea Rumph. Herb. Amb. 6: 82 (haud t. 37, 
f. 1). 

Phyllitis amboinica II terrestris Rumph. Herb. Amb. 6: 82 (haud 
t. 37, f. 2). 

Amboina, Ayer putri, Robinson PI. Rumph. Amb. 448, July 28, 1913, 
epiphytic at low altitudes. 

The descriptions given by Rumphius both apply to an Asple- 
nium of the nidus group, but perhaps more than one species is in- 
cluded. The figures are poor, and the one supposed to represent 
Phyllitis amboinica II terrestris does not agree at all with the 
plant described; it may be some species of Vittaria or Polypo- 
dium. The figure supposed to represent Phyllitis amboinica I 
arborea is almost equally poor for Asplenium nidus Linn., but 
may have been drawn from a straggling specimen. The form 
merely mentioned as having fronds much smaller and narrower 

144971 5 



66 



RUMPHIUS'S HERBARIUM AMBOINENSE 



than the above, which Hasskarl indicated as Phyllitis amboinica 
III, is indeterminable, but is certainly no Asplenium; it may be a 
Vittaria. 

BLECHNUM Linnaeus 

BLECHNUM ORIENTALE Linn. Sp. PI. (1753) 1077. 

Polypodium simplex Burm. f. PI. Ind. (1768) 235 (type!). 
Lonchitis amboinica recta I major rubra Rumph. Herb. Amb. 6: 70, 
t. 30, f. 1. 

Under the name Lonchitis amboinica Rumphius described sev- 
eral quite unrelated species, apparently pertaining to as many 
different genera as he described forms. The form figured, t. 30, 
f. I, which manifestly is recta I major rubra, is unquestionably 
Blechnum orientate Linn. It was reduced by Burman f. to 
Polypodium simplex Burm. f., a species apparently typified by 
the Rumphian illustration, and one that has remained of uncer- 
tain status until the present time. Loureiro erroneously re- 
ferred it to Pteris vittata Linn., while Henschel, following Blume, 
placed it as an undetermined species of Angiopteris. The red 
color of the young pinnae, mentioned by Rumphius, is very char- 
acteristic of Blechnum orientate Linn. 

STENOCHLAENA J. Smith 

ST ENOCH LA EN A PALUSTRIS (Burm. f.) Bedd. Ferns Brit. Ind. Suppl. 

(1876) 26. 

Polypodium palustre Burm. f. Fl. Ind. (1768) 234. 

Lonchitis amboinica III volubilis Rumph. Herb. Amb. 6: 71, t. 31. 

Not represented in our Amboina collections. The figure, how- 
ever, unquestionably represents a species of Stenochlaena, and 
from the description quite certainly S. palustris (Burm. f.) Bedd. 
It has been reduced by Willdenow to Lomaria scandens Willd., 
and by Poiret to Onoclea scandens Sw., both synonyms of Steno- 
chlaena palustris Bedd. 

CHEILANTHES Swartz 

CHEILANTHES TEN U I FOLIA (Burm. f.) Sw. Syn. (1806) 129, 332. 

Trichomanes tenuifolia Burm. f. Fl. Ind. (1768) 237. 

Acrostichum tenue Retz. Obs. 6 (1791) 39. 

Dryopteris campestris Rumph. Herb. Amb. 6: 74, t. 3J+, f. 2. 
Amboina, Way tommo, Robinson PI. Rumph. Amb. Uh2, August 19, 1913, 
terrestrial, altitude about 80 meters. 

The Rumphian illustration is unmistakably this species. It 
was first reduced by Burman f., in the original description of 
Trichomanes tenuifolia Burm f., and has been cited under the 
synonyms given above as well as under the additional synonym 
Adiantum varians Poir. 



POLYPODIACBAE 



67 



ADIANTUM Linnaeus 

ADIANTUM sp. 

Capillus veneris amboinicus Rumph. Herb. Amb. 6: 77, t. 3J>, f. 1. 

Not represented in our Amboina collections. The figure rather 
strongly resembles the Philippine Adiantum opacum Copel. Lour- 
eiro erroneously referred it to Adiantum capillus veneris Linn., 
while Pritzel placed it with equal error under Adiantum aethio- 
picum Thunb. Hasskarl, Neue Schlussel (1866) 165, suggested 
that it might be Adiantum pulchellum Blume, but if correctly 
drawn the figure represents a species quite different from the 
one described by Blume. 

ADIANTUM sp.? 

Dryopteris silvestris III petraea Rumph. Herb. Amb. 6: 74. 

Hasskarl, Neue Schlussel (1866) 165, has suggested that this 
may be Adiantum pulchellum Blume. It is probably not Blume's 
species, but is certainly an Adiantum or a Lindsaya. 

POLYPODIUM Linnaeus 

POLYPODIUM SINUOSUM Wall. Cat. (1829) no. 2231, nomen nudum; 
Hook. Sp. Pil. 5 (1863) 61, t. 284. 
Scolopendria II minor Rumph. Herb. Amb. 6: 84. 
Amboina-, Kati-kati, Robinson PI. Rumph. Amb. October 17, 1913, 

on trees in mangrove swamps. 

There is some doubt as to whether or not Scolopendria II 
minor Rumph. is Polypodium sinuosum Wall, as here interpreted. 
It seems probable that more than one species of Polypodium is 
included in Rumphius's description. 

POLYPODIUM PHYMATODES Linn. Mant. 2 (1771) 306. 

Polypodium excavatum Roxb. Hort. Beng. (1814) 75 (type!). 
Polypodium indicum II minus Rumph. Herb. Amb. 6: 80, t. 35, f. 2. 

Amboina, Ayer putri, Robinson PL Rumph. Amb. July 28, 1913, 

epiphytic at low altitudes. 

The specimen is very typical Polypodium phymatodes Linn, 
and agrees with Rumphius's description and figure. Burman 
f., Fl. Ind. (1768) 233, referred it to Polypodium dissimile Linn., 
a species of doubtful status, based on a figure of Plukenet's, 
drawn from American material. It is the type of Polypodium 
excavatum Roxb., as originally published in the Hortus Bengal- 
ensis, by citation of the Herbarium Amboinense; see C. B. 
Robinson in Philip. Journ. Sci. 7 (1912) Bot. 415. It may not, 
however, be the species actually described by Roxburgh under 
the same name in Calcutta Journ. Nat. Hist. 4 (1844) 485. 



68 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Hasskarl, Neue Schltfssel (1866) 166, suggested Polypodium 
phymatodes Linn, as the proper place for the Rumphian species, 
and I consider that this is manifestly the correct disposition 
of it. 

DRYNARIA (Bory) J. Smith 

DRYNARIA SPARSISORA (Desv.) Moore Index Fil. (1862) 348. 
Polypodium sparsisorum Desv. in Berl. Mag. 5 (1811) 315. 
Polypodium indicum 1 pilosum s. majus Rumph. Herb. Amb. 6: 78, 
t. 36. 

Amboina, Ayer putri, Robinson PL Rumph. Amb. £50, July 28, 1913, 
epiphytic at low altitudes. 

This was originally reduced by Linnaeus to Polypodium querci- 
folium Linn. (=Drynaria quercifolia J. Sm.) in Stickman Herb. 
Amb. (1754) 26, Amoen. Acad. 4 (1759) 134, Syst. ed. 10 (1759) 
1325, Sp. PL ed. 2 (1763) 1547, which has been accepted by all 
authors who have had occasion to cite the Rumphian illustra- 
tion. There is nothing in the figure by which the Rumphian 
species can be distinguished as between Drynaria sparsisora 
Moore and D. quercifolia J. Sm., the two being very closely 
allied. The Amboina specimens, however, are Drynaria sparsi- 
sora Moore, and hence the presumption is that Polypodium 
indicum I majus Rumph. pertains to this species rather than to 
Drynaria quercifolia J. Sm. 

PLATYCERIUM Desvaux 

PLATYCERIUM CORONARIUM (Koenig) Desv. Prodr. (1827) 213. 
Osmunda coronaria Koenig Naturf. Halle 21 (1785) 107, t. 3. 
Simbar majangan Rumph. Herb. Amb. 6: 83. 

This species is not represented in our Amboina collections. 
The plant described is manifestly a Platycerium, and in all 
probability it is P. coronarium Desv. Blume reduced it to 
Platycerium biforme BL, which is a synonym of P. coronarium 
Desv. 

PARKERIACEAE 

CERATOPTERIS Brongniart 

CERATOPTERIS TH ALICTROI DES (Linn.) Brongn. in Bull. Soc. Philom. 
(1821) 186. 

Acrostichum thalictroides Linn. Sp. PI. (1753) 1070. 
Acrostichum siliquosum Linn. 1. c. 

Millefolium aquaticum Rumph. Herb. Amb. 6: 176, t. 74, f. 1. 

This common and widely distributed fern is not represented in 
our Amboina collections. Millefolium aquaticum Rumph. was 



GLE IC H E NI ACE AE— SC H IZ AE ACE AE 



69 



first reduced by Linnaeus to Acrostichum siliquosum Linn., a 
synonym of Ceratopteris thalictroides (Linn.) Brongn., which is 
manifestly the correct disposition of it. It has also been cited 
under the following synonyms of Ceratopteris thalictroides 
Brongn. : Ellobocarpus oleraceus Kaulf . and Pteris thalictroides 
Willd. 

GLEICHENIACEAE 

GLEICH EN I A Smith 

GLEICH EN I A LINEARIS (Burm. f.) Clarke in Trans. Linn. Soc. Bot. 1 
(1880) 428. 

Polypodium lineare Burm. f. Fl. Ind. (1768) 235, t. 67, f. 2. 
Filix calamaria Rumph. Herb. Amb. 6: 85, t. 38. 

Amboina, Batoe merah, Robinson PI. Rumph. Amb. 446, July 18, 1913, 
on rocks at low altitudes, locally known as paku kawa. 

This reduction of Filix calamaria is certainly correct, for the 
figure and description are unmistakable. It has been reduced by 
Poiret to Polypodium dichotomum Thunb., by Willdenow to 
Mertensia dichotoma Willd., by Blume to Gleichenia hermannii 
R. Br., and by Mettenius to Gleichenia dichotoma Hook. var. 
alternans Mett. 

SCHIZAEACEAE 

SCHIZAEA Smith 

SCHIZAEA DICHOTOMA (Linn.) Smith in Mem. Acad. Turin 5 (1793) 
t. 9, f. 9. 

Acrostichum dichotomum Linn. Sp. PI. (1753) 1068. 
Equisetum silvestre III Rumph. Herb. Amb. 6: 92. 
Amboina, Salahoetoe, Robinson PI. Rumph. Amb. 460, November, 1913. 

This reduction was first made by Hasskarl, Neue Schliissel 
(1866) 168, following Blume's reduction of it to the genus 
Schizaea, and this is apparently the correct disposition of it. 

LYGODIUM Swartz 
LYGODIUM CIRCINNATUM (Burm. f.) Sw. Syn. (1806) 153. 
Ophioglossum circinnatum Burm. f. Fl. Ind. (1768) 228. 
Adianthum volubile I polypoides Rumph. Herb. Amb. 6: 75, t. S3 

(including medium and scriptum) . 
Amboina, Bin ting and Amahoesoe, Robinson PL Rumph. Amb. 451, 452, 
August, 1913, in limestone regions at low altitudes. 

Linnaeus originally reduced this, through error, to Ophioglos- 
sum flexuosum Linn., in Stickman Herb. Amb. (1754) 26, Amoen. 
Acad. 4 (1759) 134. Burman f., Fl. Ind. (1768) 228, cites it 



70 



RUMPHIUS'S HERBARIUM AMBOINENSE 



as a synonym in the original description of Ophioglossum circin- 
natum Burm. f., the species being based primarily on Javan 
specimens; it has, since Burman's species was proposed, been 
cited under this name or synonyms of it. The forms described 
as medium and scriptum are manifestly referable to Lygodium 
circinnatum Sw. 

LYGODIUM SCAN DENS (Linn.) Sw. in Schrad. Journ. 1800 2 (1801) 106. 

Ophioglossum scandens Linn. Sp. PI. (1753) 1063. 

Adianthum volubile III minus Rumph. Herb. Amb. 6: 76, t. 32, f. 2, 3. 

Amboina, Soja road and vicinity of the town of Amboina, Robinson PL 
Rumph. Amb. ^53, August and October, 1913, in thickets and forests, 
altitude 30 to 70 meters, locally known as paku kawa. 

The reduction to Ophioglossum scandens Linn, was made 
originally by Linnaeus in Stickman Herb. Amb. (1754) 26, 
Amoen. Acad. 4 (1759) 134, Syst. ed. 10 (1759) 1318, but it is 
to be noted that Linnaeus quotes "Dryopteris triplex" as the 
name corresponding to t. 32, under Ophioglossum scandens. 
The form figured is an excellent representation of Lygodium 
scandens (Linn.) Sw. Blume has referred it to Lygodium 
microphyllum R. Br., a synonym of L. scandens Sw. 

OPHIOGLOSSACEAE 

OPHIOGLOSSUM Linnaeus 

OPHIOGLOSSUM PENDULUM Linn, in Stickman Herb. Amb. (1754) 27, 
Amoen. Acad. 4 (1759) 135, Sp. PI. ed. 2 (1763) 1518 (type!). 
Scolopendria I major Rumph. Herb. Amb. 6: 84, t. 37, f. 3. 

Amboina, Soja, Robinson PL Rumph. Amb. UUO, August 2, 1913, in 
forests at an altitude of 400 meters. 

Scolopendria major Rumph. is the whole basis of Ophioglossum 
pendulum Linn., this reduction having been accepted by all 
authors, although by some placed in another genus, as Ophio- 
derma pendula (Linn.) Presl. 

OPHIOGLOSSUM PEDUNCULOSUM Desv. in Berl. Mag. 5 (1811) 306. 
Ophioglossum simplex Rumph. Herb. Amb. 6: 152, t. 88, f. 2. 

Not represented in our Amboina collections. Ophioglossum 
simplex Rumph. was originally reduced by Linnaeus to Ophio- 
glossum vulgatum Linn, in Stickman Herb. Amb. (1754) 28, 
Amoen. Acad. 4 (1759) 135, which is manifestly a wrong 
disposition of it. Roxburgh, Calcutta Journ. Nat. Hist. 4 (1844) 
475, placed it under Ophioglossum cordifolium Roxb., and 



MARATTIACEAE— LYCOPODIACEAE 



71 



Schlechtendal, Adumbr. (1825) 9, placed it under Ophioglossum 
moluccanum Schlecht., both synonyms of Ophioglossum peduncu- 
losum Desv. 

H ELM I NTHOSTACH YS Kaulfuss 

H ELM INTHOSTACH YS ZEYLANICA (Linn.) Hook. Gen. Fil. (1840) t. U7. 
Osmunda zeylanica Linn. Sp. PI. (1753) 1063. 
Ophioglossum laciniatum Rumph. Herb. Amb. 6: 153, t. 68, f. 3. 
Amboina, Kati-kati, Robinson PI. Rumph. Amb. J>55, October 28, 1913, 
in ravines at an altitude of about 70 meters. 

The original reduction of Ophioglossum laciniatum was made 
by Linnaeus (to Osmunda zeylanica Linn.) in Stickman Herb. 
Amb. (1754) 28, Amoen. Acad. 4 (1759) 135, Syst. ed. 10 (1759) 
1318", Sp. PI. ed. 2 (1763) 1519, which as Helminthostachys 
zeylanica Hook, is manifestly the correct disposition of it. By 
other authors it has been cited under Botrychium zeylanicum 
Willd., and Helminthostachys dulcis Kaulf. — both synonyms of 
H. zeylanica Hook. 

MARATTIACEAE 

ANGIOPTERIS Hoffmann 

ANGIOPTERIS AMBOINENSIS DeVr. in Nederl. Kruidk. Arch. 3 (1852) 
195, Monogr. Marat. (1853) 32. 
Filix aquatica I femina Rumph. Herb. Amb. 6: 65, t. 28. 

Not represented in our Amboina collections. Blume thought 
that this might be a species of Marattia, but the size of the plant, 
as indicated by Rumphius, makes this suggested reduction an 
impossible one. While it is impossible definitely to state that 
Filix aquatica Rumph. is identical with Angiopteris amboinensis 
DeVr., the presumption is very great that they are the same. 

LYCOPODIACEAE 

LYCOPODIUM Linnaeus 

LYCOPODIUM CERNUUM Linn. Sp. PI. (1753) 1103. 

Cingulum terrae Rumph. Herb. Amb. 6: 87, t. U0, f. 1. 

Amboina, Batoe merah and Soja road, Robinson PI. Rumph. Amb. U57, 
July and August, 1913, in rocky places and on grassy hillsides, altitude 
15 to 200 meters, locally known as daun rai rai. 

Linnaeus originally reduced Cingulum terrae to Lycopodium 
canaliculatum Linn, in Stickman Herb. Amb. (1754) 27, Amoen. 
Acad. 4 (1759) 135, but later, Syst. ed. 10 (1759) 1330, placed 
it under Lycopodium cernuum Linn., where it manifestly belongs. 



72 



RUMPHIUS'S HERBARIUM AMBOINENSE 



LYCOPODIUM PHLEGMARIA Linn. Sp. PI. (1753) 1101. 

Equisetum amboinicum s. arboreum squamatum Rumph. Herb. Amb. 

6: 91, t. U, f. 1. 

Amboina, Binting, Robinson PL Rumph. Amb. J±56, August 13, 1913, 
on trees. 

This reduction was first made by Linnaeus in Stickman Herb. 
Amb. (1754) 27, Amoen. Acad. 4 (1759) 135, Syst. ed. 10 (1759) 
1330, and is certainly the correct disposition of it. 

LYCOPODIUM NUMMULARI FOLIUM Blume Enum. (1828) 263. 
Equisetum amboinicum II minor Rumph. Herb. Amb. 6: 92? 
Not represented in our Amboina collections. The reduction is 
that suggested by van Alderwerelt van Rosenburgh in Philip. 
Journ. Sci. 11 (1916) Bot. 120. Hasskarl, Neue Schlussel 
(1866) 167, thought that it might be Lycopodium phlegmarioides 
Spring. 

SELAGINELLACEAE 

SELAGI N ELLA * Spring 

SELAGI N ELLA PLANA (Desv.) Hieron. in Engl. & Prantl Nat. Pflan- 
zenfam. 1 4 (1900) 703. 
Lycopodium planum Desv. in Lam. Encycl. Suppl. 3 (1813) 554. 
Muscus fruticescens femina Rumph. Herb. Amb. 6: 86, t. 39, f. 1. 
Amboina, Gelala, Robinson PL Rumph. Amb. 458, July 16, 1913, on 
banks at low altitudes. 

Undoubtedly this is the correct disposition of Muscus frutices- 
cens femina Rumph., and is that suggested by Desvaux in Lam. 
Encycl. Suppl. 3 (1813) 538. The only other suggested reduc- 
tion is that of Hasskarl, Neue Schlussel (1866) 167, who thought 
it might be Lycopodium dichotomum Sw. 

SELAGI N ELLA D'URVILLEI A. Br. in Verh. Zool. Bot. Ges. (1869) 585. 
Muscus fruticescens mas Rumph. Herb. Amb. 6: 86, t. 39, f. 2. 
Amboina, Hatiwe, Robinson PL Rumph. Amb. 459, September 4, 1913, 
in light woods at low altitudes. 

This is in all probability the correct disposition of Muscus 
fruticescens mas Rumph. ; although, if a number of allied species 
should be found in Amboina, it would be difficult or impossible 
to determine to which of the forms the Rumphian figure applies. 
Desvaux, in Lam. Encycl. Suppl. 3 (1813) 558, thought that it 
might be Lycopodium caudatum Desv., and Hasskarl, Neue 
Schlussel (1866) 167, thought that it might be Lycopodium 
fruticulosum Blume, both of these being species of Selaginella. 

* Retained name, Brussels Congress; Selaginoides Boehm. (1760), Lyco- 
podioides Boehm. (1760), and Stachygynandrium Beauv. (1804) are older. 



PSILOTACEAE 



73 



PSILOTACEAE 

PS I LOT UM Swartz 

PSILOTUM TRIQUETRUM Sw. Syn. (1806) 117. 

Equisetum secundum Rumph. Herb. Amb. 6: 92. 
Amboina, Amahoesoe, Bato Gadjah, and vicinity of the town of Am- 
boina, Robinson Pi. Rumph. Amb. £61, 4.62, August and September, 1913, 
on trees, sea level to an altitude of 150 meters. 

Hasskarl, Neue Schlussel (1866) 167, placed this under Psi- 
lotum complanatum Sw., but as the Amboina specimens are all 
referable to P. triquetrum Sw., it is assumed that this is the 
correct disposition of Equisetum secundum Rumph. 

PTERIDOPHYTA OF UNCERTAIN STATUS 

Filix aquatica II mas Rumph. Herb. Amb. 6: 66. 

Very briefly described in the same chapter with Angiopteris amboinensis 
DeVr. Hasskarl, Neue Schlussel (1866) 164, has referred it to Pteris 
longipes Don, but without good reason. Its status cannot be determined 
from the data given by Rumphius. 

Filix urens Rumph. Herb. Amb. 6: 69. 

This is indeterminable from any data given by Rumphius; perhaps a 
Dryopteris. 

Lonchitis amboinica recta I major alba Rumph. Herb. Amb. 6: 70. 

Van Alderwerelt van Rosenburgh has suggested that this may be Poly- 
podium albens Blume. 

Lonchitis amboinica recta II minor Rumph. Herb. Amb. 6: 70. 

Under this two forms are described, nigra and alba, both indeterminable 
from any data at present available. Willdenow, Sp. PI. 5 (1810) 228, re- 
ferred the figure, t. SO, f. 2, to Aspidium amboinense Willd., which is supposed 
to be the same as Dryopteris parasitica O. Kuntze. The figure certainly does 
not represent the latter species, and there is, moreover, no way of deter- 
mining which form Rumphius intended it to represent as between the forms 
major and minor. 

Lonchitis saguaria Rumph. Herb. Amb. 6: 72. Indeterminable. 
Lonchitis amara Rumph. Herb. Amb. 6: 72. Indeterminable. 
Lonchitis pilosa Rumph. Herb. Amb. 6: 72. Indeterminable. 
Lonchitis muscosa Rumph. Herb. Amb. 6: 72. Indeterminable. 

The above four forms are briefly described. A more comprehensive 
exploration of Amboina may yield data and material by which they can 
be eventually determined. 

Dryopteris silvestris II arborea Rumph. Herb. Amb. 6: 74. 

Indeterminable. An epiphytic fern, perhaps belonging in the Davalliae 
as suggested by Hasskarl. 



74 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Filix lanuginosa Rumph. Herb. Amb. 6: 69. 

From the description the plant must be Cibotium baranetz J. Sm. or 
Dicksonia sorbifolia Sm., as suggested by van Alderwerelt van Rosenburgh. 
A future exploration of Amboina will doubtless yield material by which its 
status can be definitely determined. 
Filix canarina Rumph. Herb. Amb. 6: 64. 

Indeterminable from the data and the material at present available. 

SPERMATOPHYTA 

GYMNOSPERMAE 
CYCADACEAE 
CYCAS Linnaeus 

CYCAS RUMPHII Miq. in Bull. Soc. Phys. Nat. Neerl. (1839) 45. 
Olus calappoides Rumph. Herb. Amb. 1: 86, t. 22, 23. 
Olus calappoides II e Celebes Rumph. 1. c. 87, t. 20, 21. 

Amboina, Wae, Robinson PI. Rumph. Amb. 563, November 29, 1913, in 
light forests near sea level, staminate, locally known as sayor kalappa. 

The specimen represents Olus calappoides mas Rumph. Herb. 
Amb. 1 : 99, t. 23. Possibly more than one species is represented 
by the plants described and figured by Rumphius, but more 
abundant material and a critical study of all the Indo-Malayan 
forms allied to Cycas circinalis Linn, will be necessary definitely 
to settle this point. The Rumphian illustrations are as follows : 
t. 20 represents an oblong-ovoid staminate inflorescence with 
leaves, poor ;t. 21, a habit sketch of the same ; t. 22, a habit sketch 
and female inflorescence and infructescence, fairly good; and 
t. 23, a habit sketch with an elongated staminate inflorescence. 

It is by no means certain that Cycas rumphii Miq. is specifically- 
distinct from C. circinalis Linn. The latter should be interpreted 
by Ceylon and Indian specimens, although in the orginal descrip- 
tion Linnaeus gives two references to Amboina figures and de- 
scriptions, including Olus calappoides Rumph. Herb. Amb. 1 : 86, 
t. 22, 23. In Stickman, Herb. Amb. (1754) 6, and in Amoen. 
Acad. 4 (1759) 119, t. 21 to 23 are included. Loureiro, Fl. 
Cochinch. (1790) 632, cites all four figures under Cycas inermis 
Lour., a species that must be interpreted from specimens from 
southern China and Cochin-China ; it is supposed to be the same 
as Cycas revoluta Thunb., but Loureiro's description of the leaves 
does not conform with Thunberg's species. Blume, Rumphia 4 
(1848) 14, refers Lagolo Rumph. Herb. Amb. 1: 87, t. 22, B, to 
Cycas thouarsii R. Br., a species of eastern Africa and Madagas- 



TAXACEAE 



75 



car and manifestly very closely allied to Cycas rumphii Miq. 
Hamilton, Mem. Wern. Soc. 5 (1826) 322, refers Olus calappoides 
Rumph., t. 20, 21, to Cycas pectinata Ham., a species described 
from Indian specimens, and one not included in Index Kewensis. 
Miquel, Comment. Phyt. (1840) 126, refers Olus calappoides 
II e Celebes to Cycas celebica Miq., a species apparently to be 
interpreted from Rumphius's description. Doctor Stapf * has 
contrasted Cycas thouarsii R. Br., C. rumphii Miq., and C. 
circinalis Linn., giving in synoptical form the macroscopic and 
microscopic characters by which the three may be distinguished. 
Pending a critical revision of the entire genus, it is probably best 
to retain the Moluccan form, that is manifestly closely allied to 
Cycas circinalis Linn., under the name Cycas rumphii Miq. 

CYCAS REVOLUTA Thunb. Fl. Jap. (1784) 229. 

Arbor calappoides sinensis Rumph. Herb. Amb. 1: 92, t. 2U> 

This commonly cultivated species is not represented in our 
Amboina collections. The Rumphian species is manifestly Cycas 
revoluta Thunb. The drawing represents a leaf only, but is 
well executed and characteristic of the species. 

TAXACEAE 

PODOCARPUSf Persoon 

PODOCARPUS RUMPHII Blume Rumphia 3 (1847) 214. 
Lignum emanum Rumph. Herb. Amb. 3: 47, t. 26. 
Amboina, Hoetoemoeri road and Hitoe messen, Robinson PI. Rumph. 
Amb. 309, September 30 and October 18, 1913, altitude 250 and 700 meters, 
locally known as dammar puti (properly the name for Agathis alba Foxw., 
p. 76). 

This specimen is Lignum emanum Rumph., but it may not be 
the same as the form on which Blume actually based his descrip- 
tion of Podocarpus rumphii. It should be critically compared 
with the species commonly known as Podocarpus neriifolius Don. 
According to Blume it is Cerbera nereifolia Zipp. in Bijdr. Nat. 
Wetensch. 5 (1830) 175, but Podocarpus neriifolius Don is earlier. 
Lignum emanum Rumph. is cited by Blume as a synonym of 
Podocarpus rumphii. 

Hasskarl, Neue Schliissel (1866) 38, refers Dammara alba mas 
Rumph. Herb. Amb. 2: 175, t. 57, f. A-C, to Podocarpus latifolia 

* Cycas Thouarsii. Kew Bull. (1916) 1-8. 

f Retained name, Vienna Code; Nageia Labill. (1806) is older. 



76 



RUMPHIUS'S HERBARIUM AMBOINENSE 



B\ume=Podocarpus blumei Endl. Of the figures cited, "A" is 
a staminate cone of Agathis alba Foxw. ; "B" is, according to 
Rumphius, a leaf of the true dammar, Agathis alba Foxw. ; while 
"C" is said by Rumphius to be a branchlet from a young tree 
of the female dammar. I can see no reason for considering that 
Podocarpus blumei Endl. is included in the description of Dam- 
mara alba Rumph. The leaves of young plants of Agathis alba 
Foxw. very greatly resemble those of Podocarpus blumei Endl., 
and it is, of course, possible that the two were confused by 
Rumphius. 

PINACEAE 

AGATHIS * Salisbury 

AGATHIS ALBA (Lam.) Foxw. in Philip. Journ. Sci. 5A (1910) 173, 6 
(1912) Bot. 167. 

Dammara alba Lam. Encycl. 2 (1786) 259 (type!). 

Pinus abies Lour. Fl. Cochinch. (1790) 579. 

Agathis loranthifolia Salisb. in Trans. Linn. Soc. 8 (1807) 311. 

Abies dammara Poir. in Lam. Encycl. Suppl. 5 (1817) 35. 

Agathis dammara Rich. Comm. Conif. Cyc. (1826) 93, t. 19. 

Pinus dammara Lamb. Pin. 1 (1803) 61, t. 38. 

Dammara rumphii Presl Epim. Bot. (1851) 236. 

Dammara alba Rumph. Herb. Amb. 2: 174, t. 57. 

Amboina, Soija diatas, Robinson PL Rumph. Amb. 220, locally known 
as dammar puti. 

Dammara alba Rumph. is the whole basis of Dammara alba 
Lam. Lamarck published the species with Rumphius as authority. 
It must be interpreted from the Rumphian figure and description 
and from Amboina specimens, while most of the synonyms 
cited above must be interpreted wholly or partly from the same 
data. Warburg, Monsunia 1 (1900) 182, in his attempt to 
split up the collective species Agathis dammara (Lamb.) Rich, 
states : "Diese Art kann nur nach der Beschreibung von Rumph 
rekonstruirt werden, da sicheres aus Amboina stammendes Mat- 
erial leider nicht vorliegt und die vielfachen in der Litteratur 
gegebenen Beschreibungen der Sammeltart nach dem verschie- 
densten Material angefertigt sind." Warburg recognized ten 
species of Agathis from the Malayan region, but the status of 
these as valid ones is open to grave doubt. 

Dammara alba mas Rumph. Herb. Amb. 2: 175, t. 57, f. A-C, referred 
by Hasskarl, Neue Schiissel (1866) 38, to Podocarpus latifolia Blume, 
I consider to be Agathis alba (Lam.) Foxw. 



* Retained name, Vienna Code; Dammara Lam. (1786) is older. 



GNETACEAE 



77 



Dammara alba femina Rumph. 1. c. 175, t. 57, f. D is surely Agathis 
alba (Lam.) Foxw. 

Dammara celebica Rumph. Herb. Amb. 2: 179 = Dammara alba var. 
celebica Hassk., Neue Schlussel (1866) 38, is suggested by Warburg to 
be the same as Agathis celebica (Koord.) Warb. Monsunia 1 (1900) 185, 
but Doctor Foxworthy, who has examined Celebes specimens collected by 
Koorders, considers the species to be a synonym of Agathis alba Foxw. 
and reduces likewise Agathis borneensis Warb., A. beccarii Warb., and 
A. macrostachys Warb. 

GNETACEAE 
GNETUM Linnaeus 

GNETUM GNEMON Linn. Mant. 1 (1767) 125. 

Gnetum ovalifolium Poir. in Lam. Encycl. Suppl. 2 (1811) 810. 

Gnemon domestica mas Rumph. Herb. Amb. 1 : 181, t. 72. 

Gnemon domestica femina Rumph. Herb. Amb. 1:181, t. 71. 

Gnemon silvestris Rumph. Herb. Amb. 1: 183, t. 73. 
Amboina, Soja, Robinson PI. Rumph. Amb. 213, August 31, 1913, in 
forests at 300 meters altitude (Gnemon silvestris Rumph.) ; Robinson PI. 
Rumph. Amb. 209, October 24, 1913, in light woods, altitude about 250 
meters, both locally known as gnemo. 

I consider that the three forms described and figured by 
Rumphius represent but a single species, this being practically 
the opinion of other authors who have considered the status of 
the Rumphian species. Gnemon domestica femina Rumph. was 
reduced by Linnaeus to Gnetum gnemon Linn, in the original 
description of the species, the only deviation from this reduction 
presented in botanical literature being Blume's reference of it 
to Gnetum gnemon Linn. var. laurinum Blume Rumphia 4 (1848) 
3, together with Gnemon domestica mas Rumph. Gnetum oval- 
ifolium Poir. was based on specimens collected in Amboina by 
Labillardiere, with an added reference to Gnemon silvestris 
Rumph. Blume has reduced it to Gnetum gnemon Linn, as var. 
ovalifolium (Poir.) Blume, but I consider it scarcely distinguish- 
able from typical Gnetum gnemon Linn, even as a variety. 
Gnemon domestica mas Rumph. was reduced by Blume to Gnetum 
gnemon Linn., var. lucidum Blume, Rumphia 4 (1848) 4. 

GNETUM INDICUM (Lour.) comb. nov. 

Abutua indica Lour. Fl. Cochinch. (1790) 630. 

Gnetum funicular e Brongn. in Duperry Voy. Bot. (1829) 12. 

Gnetum funicular e Blume Nov. Fam. (1834) 32, Hoev. & DeVriese 

Tijdschr. 1 (1834) 162, Ann. Sci. Nat. II 2 (1834) 106. 
Gnetum latifolium Blume op. cit. 30, 162, 105. 
Gnemon funicularis Rumph. Herb. Amb. 5: 12, t. 8. 



78 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Not represented in our Amboina collections. Loureiro quotes 

Gnemon funicularis Rumph. as a synonym of Abutua indica Lour, 
in the original description of that species. Louriero's type, 
in the herbarium of the British Museum, is a leaf specimen, 
and according to Doctor Rendie, who has examined it for me, is 
apparently the same as Gnetum funiculare Blume. Roxburgh, 
Hort. Beng. (1814) 66, based his Gnetum scandens on "H. M. 
7. t. 22; H. A. 5. t. 7, 8," i. e. the first reference to Rheede 
Hortus Malabaricus, and the second to Rumphius Herbarium 
Amboinense ; I believe that the species should be typified by the 
first reference. However, "Ula. Rheed. mal. 7. p. 41. t. 22" 
is the whole basis of Thoa edulis Willd. Sp. PI. 4 (1805) 477, 
so that Gnetum scandens Roxb. becomes a synonym of Gnetum 
edule (Willd.) Blume, together with Gnetum ula Brongn. The 
plant that Blume actually described as Gnetum edule seems 
not to be the same as the Indian Thoa edulis Willd., but the 
name must go with the Indian plant. It is by no means clear 
that this continental form, which appears in modern literature 
as Gnetum scandens Roxb., is specifically distinct from the 
Malayan Gnetum indicum (Lour.) Merr., but at any rate, 
Loureiro's specific name is much older than any of the others. 

GNETUM GNEMONOIDES Brongn. in Duperry Voy. Bot. (1829) 12 
(type!). 

Gnetum rumphianum Becc. Malesia 1 (1877) 182. 
Gnetum verrucosum Karst. in Ann. Jard. Bot. Buitenz. 11 (1893) 
216. 

Funis gnemoniformls Rumph. Herb. Amb. 5: 12, t. 8. 

Not represented in our Amboina collections. Gnetum gne- 
monoides Brongn. was based wholly on the Rumphian figure 
and description of Funis gnemoniformis. Blume, Fam. Nov. 
(1834) 31, Ann. Sci. Nat. II 2 (1834) 106, reduced Funis 
gnemoniforw.is Rumph. to Gnetum edule Blume, a species based 
on Thoa edulis Willd. and differing remarkably from Rumphius's 
species, as described, in its fruit characters. Gnetum rumphia- 
num Becc. was based on specimens from New Guinea, with 
the addition of a reference to Funis gnemoniformis Rumph. 
It has seeds 5 to 5.5 cm long, in entire agreement with 
Rumphius's description of the fruits of Funis gnemoniformis 
as "tres digitos transversales longi." Gnetum verrucosum 
Karst. was described from specimens originating in Buru, with 
fruits 4.5 cm in length, and I have no hesitation in reducing it to 
Gnetum gnemonoides Brongn. 



PANDANACEAE 



79 



ANGIOSPERMAE 

( MONOCOTYLEDONS ) 

PANDANACEAE* 

PAN DAN US Linnaeus 

PANDANUS POLYCEPHALUS Lam. Encycl. 1 (1785) 372 (type!). 
Pandanus humilis Rumph. Herb. Amb. 4: 143, t. 76. 
Amboina, Binting and Lateri, Robinson PL Rumph. Amb. 5U, July and 
August, 1913, in shaded places along streams at low altitudes, and in 
forests at an altitude of about 250 meters, locally known as keker and 
pandan keker ayer. 

Pandanus humilis Rumph. is the whole basis of Pandanus 
polycephalus Lam., Lamarck's species being based wholly on 
Rumphius's figure and description. Loureiro, Fl. Cochinch. 
(1790) 603, described Pandanus humilis Lour, from Cochin- 
China material and reduced to it Pandanus humilis Rumph. 
There is every reason to suppose that the Cochin-China plant 
described by Loureiro represents a species entirely different 
from that described by Rumphius, and that Warburg was in 
error in reducing Pandanus humilis Lour, to Pandanus poly- 
cephalus Lam. The type of Loureiro's species is manifestly 
the Cochin-China plant described, not the Rumphian synonym 
cited. 

PANDANUS ROBINSON 1 1 sp. nov. § Keura. 

Pandanus spurius Rumph. Herb. Amb. 4: 142, t. 75. 
Amboina, Paso, Robinson PI. Rumph. Amb. 30 (type), October 29, 1913, 
along the seashore, locally known as keker laut. 

Arbor circiter 3.5 m alta, ramosa. Foliis circiter 1 m longis, 
4.5 cm latis, aculeatis, apice longissime tenuiterque caudatis. 
Capitulis ellipticis, 10 cm longis vel maturis majoribus, solita- 
riis, subpendulis; drupis numerosis, obconicis, circiter 2.5 cm 
longis vel paullo longioribus, apice 1.5 ad 2.5 cm latis, deorsum 
angustatis, apice latis, subtruncatis, 5- ad 10-locularis, loculis 
apice sulcis circiter 4 mm longis separatis, lobis oblique pyra- 
midatis, acutis, brunneis, nitidis, 5 ad 8 mm diametro. 

Pandanus spurius Rumph. has been referred to many different 

* I am under obligations to Dr. U. Martelli, Florence, Italy, for deter- 
minations of the Pandanaceae. For the proposed changes in nomenclature, 
however, and for the discussions of the Rumphian species, I am wholly 
responsible. 



80 



RUMPHIUS'S HERBARIUM AMBOINENSE 



species under the binomial system, but none of the names are 
valid, for which reason I propose to call this plant Pandanus 
robinsonii. Doctor Martelli, who has made the determination 
of the specimen as Pandanus spurius, would call it Pandanus 
spurius Mart., non Miq., but I consider the specific name spurius 
to be invalid. Loureiro, Fl. Cochinch. (1790) 603, erroneously- 
referred it to Pandanus odoratissimus Linn. f. Lamarck, 
Encycl. 1 (1785) 372, placed it under Pandanus odoratissimus 
as var. p. Henschel, Vita Rumph. (1833) 166, referred the 
name Pandanus spurius, but not the figure, to Pandanus fas- 
cicularis Lam., erroneously citing plates 80 and 81 as P. spurius. 
Persoon, Syn. 2 (1807) 597, placed it, with doubt, under Pan- 
danus candelabrum Beauv., an African species. Hasskarl, 
Flora 25 (1842) Beibl. 2: 14, referred it to Marquatia globosa 
Hassk., a new genus and species based on specimens cultivated 
in the botanic garden at Buitenzorg, Java, originating in 
Mauritius; this Walpers, Ann. 1 (1849) 753, renamed Hass- 
karlia globosa Walp. Both names are synonyms of Pandanus 
utilis Bory. Miquel, Anal. Bot. Ind. (1851) 57, Fl. Ind. Bat. 
3 (1855) 157, recognizing the fact that Pandanus spurius 
Rumph. was not the same as Marquatia globosa Hassk. (Hass- 
karlia globosa Walp.), adopted the specific name Pandanus 
spurius for the species, after Rumphius, but his description 
applies to the species Hasskarl described, and thus it becomes a 
synonym of Pandanus utilis Bory. 

PANDANUS REPENS Miq. Fl. Ind. Bat. 3 (1855) 165 (type!). 
Pandanus repens Rumph. Herb. Amb. 4: 152. 

Not represented in our Amboina collections. From the de- 
scription the plant is probably a variety of Pandanus tectorius 
Soland., corresponding to the Philippine form, Pandanus sabotan 
Blanco, used for similar purposes, that is, for making mats. 
A species of very doubtful status, known only from Rumphius's 
description. 

PANDANUS H ASSKARLI I nom. nov. 

Pandanus latifolius Hassk. in Flora 25 (1842) Beibl. 2: 13, non Perr., 
nec aliorum. 

Pandanus latifolius Rumph. Herb. Amb. 4: 146, t. 78. 

Not represented in our Amboina collections. There is no 
doubt whatever that Hasskarl correctly interpreted Pandanus 
latifolius Rumph. I have, however, proposed a new name for 
the species, as the name latifolius is preoccupied in the genus. 



PANDANACEAE 



81 



PANDANUS TECTORIUS Soland. in Parkins. Voy. H. M. S. Endeavour 
(1773) 46. 

Pandanus odoratissimus Linn. f. Suppl. (1781) 424. 
Pandanus verus Rumph. Herb. Amb. 4: 139, t. 7-4. 
Folium baggea Rumph. quoad t. 31. 

This common coastal species is not represented in our Amboina 
collections, yet the reduction, originally made by the younger 
Linnaeus, is undoubtedly correct ; it is to be noted, however, that 
he reduced to this species H t. 7U ad 81," of which the first and 
the last are apparently Pandanus tectorius Soland. (P. odoratis- 
simus Linn, f.), but the others represent entirely different 
species. Table 7U is exceedingly poor, but the description of 
Pandanus verus Rumph. applies unmistakably to Pandanus 
tectorius Soland. 

PANDANUS TECTORIUS Soland. var. MOSCHATUS (Miq.). 

Pandanus tectorius Soland. var. laevis (Kunth) Warb. in Engl. Pflan- 

zenreich 3 (1900) 48. 
Pandanus laevis Kunth Enum. 3 (1841) 100 (type!), non Lour. 
Pandanus moschatus Miq. Fl. Ind. Bat. 3 (1855) 165 (type!). 
Pandanus moschatus Rumph. Herb. Amb. 4: 147. 

I have followed Warburg in the reduction of this Rumphian 
species, but differ from him in the selection of the varietal 
name, as I consider Pandanus laevis Kunth to be invalidated by 
P. laevis Lour. Loureiro reduced Pandanus moschatus Rumph. 
to Pandanus laevis Lour., Fl. Cochinch. (1790) 604, but 
Loureiro's species is manifestly not the same as the one 
Rumphius described and is to be interpreted by Cochin-China 
material. All early authors, however, followed Loureiro in this 
reduction. 

PANDANUS CONOIDEUS Lam. Encycl. 1 (1785) 372 (type!). 
Pandanus ceramicus Kunth Enum. 3 (1841) 98 (type!). 
Pandanus ceramicus Rumph. Herb. Amb. 4: 149, t. 79. 

This species was reported by Rumphius from Ceram, Buru, 
Gilolo, and Ternate, but not from Amboina except as an intro- 
duced and rarely cultivated plant. It is not represented in our 
Amboina collections. The Rumphian figure and description are 
the whole basis of both Pandanus conoideus Lam. and P. 
ceramicus Kunth; and Warburg, in Engl. Pflanzenreich. 3 (1900) 
69, has apparently interpreted the species correctly. 

PANDANUS DUBIUS Spreng. Syst. 3 (1826) 897 (type!). 

Folium baggea maritimum Rumph. Herb. Amb. 4: 151, t. 80 (non 
t. 81!). 

Amboina, Latoehalat, Robinson PI. Rumph. Amb. 55, September 22, 
1913, along* the seashore, locally known as haun. 

144971 6 



82 



RUMPHIUS'S HERBARIUM AMBOINENSE 



This very characteristic species, as I interpret the original 
description by Sprengel, is typified by the Rumphian description 
and illustration, the reference to the Mascarene Islands being 
added because of the doubtful reduction of Pandanus erigens 
Thouars (=Pandanus montanus Bory). The figure of the 
single drupe given by Rumphius leaves absolutely no doubt as 
to the identity of Pandanus baggea maritimum. 

PANDANUS BAGEA Miq. PI. Ind. Bat. 3 (1855) 159 (type!). 
Folium baggea verum Rumph. Herb. Amb. 4: 150. 

This form was reduced by Warburg, in Engl. Pflanzenreich 3 
(1900) 50, as a synonym of Pandanus dubius Spreng., which is 
apparently wrong. According to the description and habitat 
given by Rumphius, it cannot possibly be Sprengel's species. It 
is suspected that it may be a form of Pandanus tectorius Soland., 
and it may be the form of Pandanus baggea Rumph. figured on 
t. 81, which I have referred to Pandanus tectorius Soland. 

PANDANUS AM BOI N ENSIS Warb. in Engl. Pflanzenreich 3 (1900) 83. 
Pandanus rumphii Warb. in Engl. Pflanzenreich 3 (1900) 84 (type!). 
Pandanus montanus Miq. PI. Ind. Bat. 3 (1855) 161 (type!), non 
Bory. 

Pandanus cer amicus Kunth var. sylvestris Kunth Enum. 3 (1841) 
98 (type!). 

Pandanus silvestris (terrestris 11) Rumph. Herb. Amb. 4:145, t. 77. 

Amboina, Lateri, Robinson PI. Rumph. Amb. 31, September 9, 1913, in 
forests, altitude about 350 meters, locally known as keker saun. 

Under Pandanus silvestris Rumphius described two entirely 
different species, but his illustration manifestly belongs with 
Pandanus silvestris terrestris II, which is the type of Pandanus 
montanus Miq., P. rumphii Warb., and the variety sylvestris of 
Pandanus ceramicus Kunth. The mature cones are about 40 
cm long and 8 to 9 cm in diameter. The species is known only 
from Amboina. 

PANDANUS TERRESTRIS Warb. in Engl. Pflanzenreich 3 (1900) 84 
(type!), excl. syn. Miquel. 
Pandanus sylvestris Miq. Fl. Ind. Bat. 3 (1855) 161 (type!), non 
Bory. 

Pandanus montanus (silvestris I) Rumph. Herb. Amb. 4: 145 (non 

t. 771). 

Anassa silvestris Rumph. Herb. Amb. 5: 230? 

Not represented in our Amboina collections, and a species of 
very doubtful status. Warburg, in Engl. Pflanzenreich 3 (1900) 
84, has erroneously cited the Miquelian synonym, Pandanus mon- 
tanus, under Pandanus terrestris Warb., but it properly belongs 
with Pandanus amboinensis Warb. (P. rumphii Warb.) ; the 



PANDANACEAE 



83 



Rumphian figure cannot belong to Pandanus terrestris Warb. 
according to Rumphius's description, but certainly belongs with 
Pandanus amboinensis Warb. Hasskarl, Neue Schlussel (1866^ 
87, has interchanged most of the synonyms cited by him between 
terrestris I and terrestris II 9 citing the plate, with doubt, under 
both. The brief description given by Rumphius is the whole 
basis of Pandanus sylvestris Miq., non Bory, and P. terrestris 
Warb. The form mentioned by Rumphius as Anassa silvestris 
is unquestionably a Pandanus and is probably referable here. 
No data are given, however, by which its exact status can be 
determined. 

FREYCI N ETI A Gaudichaud 

FREYCI N ETI A FUNICULARIS (Savigny) comb. nov. 

Pandanus funicularis Savigny in Lam. Encycl. 4 (1798) 735 (type!). 

Freycinetia strobilacea Blume Rumphia 1 (1835) 156. 

Pandanus funicularis Rumph. Herb. Amb. 4: 153 t. 82. 
Amboina, Way tommo, Robinson PI. Rumph. Amb. 29, August 16, 1913, 
in forests, altitude about 30 meters, locally known as anapur. 

Savigny compiled a description of the species, under the 
Rumphian binomial, in Lamarck's Encyclopedie 4 (1798) 735, 
and this manifestly constitutes a valid post-Linnean publication 
of the binomial. I have accordingly accepted this specific name 
in place of Blume's. Blume reduced Pandanus funicularis 
Rumph. to Freycinetia strobilacea Blume in the original descrip- 
tion of that species, the type being from Amboina. The species 
is known only from Amboina. 

FREYCINETIA GRAM I N EA Blume Rumphia 1 (1835) 159 (type!). 
Carex arborea Rumph. Herb. Amb. 6: 21, t. 8, /. 2. 

Not represented in our Amboina collections unless the sterile 
Rel. Robins. 160 J* is referable here. Carex arborea Rumph. 
is the whole basis of Freycinetia gmminea Blume and is ap- 
parently known only from the Rumphian description. Linnaeus, 
Syst. ed. 10 (1759) 865, erroneously referred t. 8, f. 2, to Schoe- 
nus secans Linn. =Scleria; but the Rumphian reference is not 
the type of the species, and the figure intended by Linnaeus was 
manifestly t. 8, f. 1, which is a Scleria. 

FREYCINETIA sp. 

Adpendix cuscuaria I angustifolia Rumph. Herb. Amb. 5: 488. 

The form described is manifestly some species of Freycinetia. 
This reduction was suggested by Hasskarl, with doubt, Neue 
Schlussel (1866) 150. Its further identification is impossible 
from the meager data given by Rumphius. 



84 



RUMPHIUS'S HERBARIUM AMBOINENSE 



HYDROCHARITACEAE 

EN H ALUS Richard 

EN H ALUS ACOROIDES (Linn, f.) Rich, ex Steud. Nomencl. ed. 2, 1 
(1840) 554; Chatin Anat. PI. Aquat. (1862) 15, t. 6. 
Stratiotes acoroides Linn. f. Suppl. (1781) 268. 
Enhalus koenigii Rich, in Mem. Inst. Paris 2 (1811) 78. 
Acorus marinus Rumph. Herb. Amb. 6: 191, t. 75, f. 2. 
Amboina, Gelala, Robinson PI. Rumph. Amb. 474, on tidal flats, October 
20, 1913. 

Acorus marinus Rumph. is the first reference given by the 
younger Linnaeus in the orginal publication of Stratiotes aco- 
roides Linn. f. ; the actual type, however, was a specimen from 
Ceylon, collected by Koenig. This reduction was followed by 
Willdenow, Poiret, Persoon, and other authors. Miquel, Fl. 
Ind. Bat. 3 (1857) 237, cites it under Enhalus koenigii Rich., 
a synonym of E. acoroides (Linn, f.) Rich. 

GRAMINEAE 

The comparatively few species of this family considered by 
Rumphius are chiefly those of economic value, such as the coarser 
forms, the bamboos, etc. As is to be expected in such a difficult 
group, it is by no means easy to determine the status? of some 
of the forms considered, this being especially true of the bamboos. 
Unfortunately numerous species of Bambusa have been based 
wholly on the descriptions or figures given by Rumphius, and 
these must be interpreted by the data given by him. Until more 
comprehensive collections are made in the Moluccas with special 
reference to the descriptions and native names given by 
Rumphius, the exact status of several of these species must 
remain doubtful. 

ZEA Linnaeus 
ZEA MAYS Linn. Sp. PI. (1753) 971. 

Frumentum indicum s. turcicum s. saracenicum Rumph. Herb. Amb. 
5: 202. 

This is merely mentioned by Burman in a note appended to 
the description of Panicum indicum s. botton, i. e. Setaria italica 
(Linn.) Kunth. It is cultivated in all parts of the Malay 
Archipelago. 

CO IX Linnaeus 
COIX LACHRYMA JOBI Linn. Sp. PI. (1753) 972. 

Lachryma jobi indica Rumph. Herb. Amb. 5: 193, t. 75. 
Lithospermum amboinicum Rumph. Herb. Amb. 6: 22, t. 9, f. 1. 

Amboina, Way tommo, along the banks of a stream, Robinson PL 
Rumph. Amb. 40, August 19, 1913, locally known as bull buli. 



GRAMINEAE 



85 



The plant figured in Volume V of the Herbarium Amboinense 
was reduced by Linnaeus to Coix lachryma jobi in Stickman 
Herb. Amb. (1754) 20, Amoen. Acad. 4 (1759) 130, and in the 
second edition of his Species Plantarum (1763) 1378, but by 
error Linnaeus cites Ova piscium of Rumphius, rather than 
Lachryma jobi indica; the two are figured on the same plate, 
and the former is referred by Hasskarl to his Saccharum edule. 
The two figures given by Rumphius represent fairly good habit 
sketches of the common Job's tears, Coix lachryma jobi Linn. 
Lithospermum amboinicum was also reduced by Linnaeus, op. 
cit. 25, 134, to Coix lachryma jobi, but by error t. 8 is cited 
instead of t. 9, although the Rumphian name is correctly given. 

IMPERATA Cyrillo 

IMPERATA CYLINDRICA (Linn.) Beauv. Agrost. (1812) expl. pi. 5, 
t. 5, f. 1, var. KOENIGII (Retz.) Benth. ex Pilger in Perk. Frag. 
Fl. Philip. (1904) 137. 
Saccharum koenigii Retz. Obs. 5 (1789) 16. 
Gramen caricosum Rumph. Herb. Amb. 6: 17, t. 7, f. 2A. 
Amboina, Hoetoemoeri road, on barren hillsides, Robinson PI. Rumph. 
Amb. 39, September 30, 1913. 

Gramen caricosum was referred by Linnaeus, Syst. ed. 10 
(1759) 869, to his Saccharum spicatum, but later, Sp. PL ed. 
2 (1763) 1480, to his Andropogon caricosus. The latter species, 
however, was based on Indian specimens and is a true Andro- 
pogon. Although the specific name was taken from Rumphius, 
the Rumphian figure and description cannot be interpreted as 
the type. Saccharum spicatum Linn, is in itself a mixture, 
but probably should be interpreted as an Imperata, not as Perotis 
latifolia Ait. The erroneous reference of Gramen caricosum to 
Andropogon caricosus by Linnaeus was followed by Burman f., 
Fl. Ind. (1768) 218; by Lamarck, Encycl. 1 (1785) 373; and by 
Willdenow, Sp. PI. 4 2 (1805) 902. Loureiro, Fl. Cochinch. 
(1790) 53, follows the first Linnean reduction and considers it 
under Saccharum spicatum Linn., while Roxburgh, Fl. Ind. ed. 2, 
1 (1832) 234, places it under Saccharum cylindricum Linn. 
=Imperata cylindrica Beauv. 

MISCANTHUS Andersson 

MISCANTHUS SINENSIS Anders, in Oefv. Vet. Akad. Forhandl. Stockh. 
(1855) 166. 
Arundo farcta I Rumph. Herb. Amb. 4: 21. 

Arundo farcta Rumph. was referred by Linnaeus, with doubt, 
to Andropogon nardus Linn., Mant. 2 (1771) 500, in which dis- 



86 



RUMPHIUS'S HERBARIUM AMBOINENSE 



position of it he was followed by Lamarck, Encycl. 1 (1785) 
374. Burman f., Fl. Ind. (1768) 30, however, placed it under 
Lagurus paniculatus Linn., which species is there properly 
published, and which is not included in Index Kewensis; it is, 
however, a synonym of Andropogon nardus Linn. The Rumph- 
ian plant, however, has nothing to do with Andropogon nardus, 
but undoubtedly is a Miscanthus. 

Ml SCAN THUS JAPONICUS (Thunb.) Anders, in Oefv. Vet. Akad. For- 
handl. Stockh. (1855) 166. 
Saccharum japonicum Thunb. in Trans. Linn. Soc. 2 (1794) 328. 
Arundo farcta II Rumph. Herb. Amb. 4: 21, t. 6. 
Amboina, Hoenoet, on dry hills, altitude about 50 meters, Robinson PL 
Rumph. Amb. 38, October 7, 1913. 

The description is short and imperfect, but apparently applies 
to this species, which, as currently interpreted, is of very wide 
distribution in eastern Asia and Malaya. It is possible that 
the species intended by the Rumphian description is Miscanthus 
floridulus (Labill.) Warb., which Warburg considers to be speci- 
fically distinct from the northern form, typical Miscanthus 
japonicus Anders. Rumphius describes the plant as from 10 
to 12 feet high, while Arundo farcta I is described as smaller. 

SACCHARUM Linnaeus 

SACCHARUM OFFICINARUM Linn. Sp. PL (1753) 54. 

Arundo saccharifera Rumph. Herb. Amb. 5: 186, t. 7U, f. 1, 2. 
Ova piscium Rumph. Herb. Amb. 5: 191, t. 75, f. 1? 

The common sugar cane is not represented in our Amboina 
collections. Three or four distinct varieties are considered by 
Rumphius, under such names as alba, fusca, rotanga, etc. The 
reduction of Arundo saccharifera Rumph. to Saccharum offici- 
narum was made by Linnaeus in Stickman's Herb. Amb. (1754) 
20, Amoen. Acad. 4 (1759) 130, Syst. ed. 10 (1759) 869, Sp. 
PL ed. 2 (1762) 79, followed by various authors. Hasskarl, 
Neue Schlussel (1866) 110, has carried the reduction of the 
various forms described by Rumphius to varieties, and considers 
Arundo saccharifera III, tabu rottang (expl. pi.) to represent 
Saccharum sinense Roxb. However, Hackel, apparently cor- 
rectly, reduces Saccharum sinense Roxb. to S. officinarum Linn. 
Ova piscium Rumph., referred by Hasskarl to Saccharum edule 
Hassk., is probably a form of Saccharum officinarum Linn. It 
was, by error, referred by Linnaeus to Coix lachryma jobi L., 
but Linnaeus manifestly intended figure 2 of plate 75, rather 
than figure 1. 



GRAMINEAE 



87 



ISCHAEMUM Linnaeus 

ISCH AE M U M TIMORENSE Kunth Rev. Gram. 1 (1835) 369, t. 98. 

Hippogrostis amboinica I major Rumph. Herb. Amb. 6: 14, t. 5, f. 2? 
Amboina, Batoe mera, along ditches, Robinson PL Rumph. Amb. U7, 
July 20, 1913. 

This figure has been confused by some authors with Panicum 
colonum Linn., and Hasskarl, Neue Schltissel (1866) 153, so 
refers it. I am not quite certain that it is referable to Ischae- 
mum timorense Kunth., but it surely is not Panicum colonum 
Linn., although Hasskarl was apparently satisfied with this 
reference of it. 

ANDROPOGON Linnaeus 

ANDROPOGON ACICULATUS Retz. Obs. 5 (1789) 22. 
Rhaphis trivialis Lour. Fl. Cochinch. (1790) 553. 
Gramen aciculatum Rumph. Herb. Amb. 6: 13, t. 5, f. 1. 
Amboina, Amahoesoe, along roadsides, Robinson PL Rumph. Amb. £5, 
August 13, 1913, locally known as rumput gintang. 

Linnaeus, Species Plantarum ed. 2 (1762) 84, erroneously 
referred Gramen aciculatum to Panicum colonum Linn., but 
while citing the name Gramen aciculatum he gives the figure 
as t. 5, f. 3, which is apparently an Oplismenus. Loureiro cites 
Gramen aciculatum in his description of Rhaphis trivialis, Fl. 
Cochinch. (1790) 553; Willdenow, Sp. PL 1 (1797) 338, repeats 
Linnaeus's error in referring it to Panicum colonum, citing the 
Rumphian name, but t. 5, f. 3, as does Linnaeus, but later, op. cit. 
4 2 (1805) 906, refers Gramen aciculatum t. 5, f. 1, to Andro- 
pogon acicularis Willd.=A. aciculatus Retz. Rumphius's figure 
is an excellent one. 

ANDROPOGON SORGHUM (Linn.) Brot. Fl. Lusit. (1804) 89, var. 
Holcus sorghum Linn. Sp. PI. (1753) 1047. 

Sorghum s. Battari Rumph. Herb. Amb. 5: 194, t. 75 bis, fig. 1. 

This is not represented in our Amboina collections. The form 
described and crudely figured by Rumphius is apparently the 
one described by Linnaeus as Holcus saccharatus=Andropogon 
sorghum var. saccharatus Hack. It was reduced to Holcus sac- 
charatus by Linnaeus in Stickman Herb. Amb. (1754) 20, fol- 
lowed in Amoen. Acad. 4 (1759) 130 (plate cited as 7U by error), 
Syst. ed. 10 (1759) 1305, followed by Loureiro, FL Cochinch. 
(1790) 645, and Willdenow, Sp. PL 4 2 (1805) 930. Burman f., 
however, FL Ind. (1768) 220, refers it to Holcus sorghum Linn., 
in which he is followed by Lamarck, Encycl. 3 (1789) 140. 
Following Hackers classification, it is probably best placed under 



88 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Andropogon sorghum Brot., subsp. sativus Hack. var. saccha- 
ratus (L.) Hack. Following other authors, it is considered 
under the generic name Sorghum, while Hitchcock proposes to 
recognize the genus Holcus for the sorghums, and retains the 
Linnean names, Holcus sorghum, H. saccharatus, etc. ; Holcus 
of authors, as typified by the European Holcus lanatus Linn., 
becomes Nothoholcus. 

ANDROPOGON AMBOINICUS (Linn.) comb. nov. 

Poa amboinica Linn. Mant. 2 (1771) 557 (type!). 
Poa amboinensis Murr. in Linn. Syst. ed. 13 (1774) 98 (type!). 
Eragrostis amboinensis Trin. ex Steud. Nomencl. ed. 2, 1 (1840) 562 
(type!). 

Phoenix amboinica montana Rumph. Herb. Amb. 6: 19, t. 7, f. 3. 

Amboina, Soja road, on grassy hillsides, altitude 300 meters, Robinson 
PI. Rumph. Amb. US, August 1, 1913. 

The specimen agrees perfectly with the description given by 
Rumphius, and sufficiently well with the figure, which is rather 
crude. The species has not been previously recognized, and Poa 
amboinica Linn.=Poa amboinensis Murr. is reduced in Index 
Kewensis to Eragrostis amboinensis Trin., this being merely a 
transfer of the specific name by Steudel. Poa amboinica Linn, 
is based wholly on Rumphius's description and figure from which 
the species must be interpreted. 

Andropogon amboinicus (Linn.) Merr., if interpreted in a 
broad sense, is identical with Andropogon serratus Thunb. ; and, 
if Hackel be followed in considering Thunberg's species as in- 
cluding several varieties, the Linnean specific name will replace 
Thunberg's. However, Andropogon amboinicus is apparently 
the form mentioned by Hackel under Andropogon serratus 
Thunb. var. genuinus Hack, subvar. major Hack, in DC. Monog. 
Phan. 6 (1889) 521. The Amboina specimen differs radically 
from typical Andropogon serratus Thunb. in its long-pilose 
sheaths and larger spikelets and probably should be considered 
specifically distinct. 

ANDROPOGON CITRATUS DC. Cat. Hort. Monsp. (1813) 78; Nees in 
Allgem. Gartenzeit. 3 (1835) 266. 
Cymbopogon citratus Stapf in Kew Bull. (1906) 322, 357, plate. 
Schoenanthemum ambolnicum Rumph. Herb. Amb. 5: 181, t. 72, f. 2. 

This species is not represented in our Amboina collections. 
For a very full discussion of it the student is referred to Doctor 
Stapf 's article.* Schoenanthemum amboinicum has very gen- 
erally been confused with Andropogon schoenanthus Linn.; in 



* The oil-grasses of India and Ceylon. Kew Bull. (1906) 297-364. 



GRAMINEAE 



89 



fact one year after Andropogon schoenanthus was published, 
Linnaeus himself referred to it the Rumphian figure in Stick- 
man's dissertation on the Herbarium Amboinense (1754) 20, 
which was repeated in Amoen. Acad. 4 (1759) 130, Syst. ed. 
10 (1759) 1304; in Murray's edition of the Systema Vegeta- 
bilium (1774) 758; by Loureiro, Fl. Cochinch. (1790) 646; 
and by Burman f., Fl. Ind. (1768) 219; the last author also in 
the same work, page 24, erroneously referring it to Panicum 
polystachion Linn. 

ANDROPOGON EXALTATUS R. Br. Prodr. (1810) 202. 

Schoenanthemum alterum Rumph. Herb. Amb. 5: 182? 
Amboina, Silali, on barren hills, altitude 125 meters, Robinson PL 
Rumph. Amb. £6, September 22, 1913, locally known as kusu kusu. 

Hasskarl has suggested Andropogon circinnatus Hochst. as 
the possible place for this form briefly mentioned by Rumphius. 
As what I take to be a form of Andropogon exaltatus R. Br. 
occurs in Amboina and as this has fragrant leaves, I merely 
make the suggestion that it may be the species intended by 
Rumphius. The species has been previously reported only from 
Australia and from Thursday Island. 

THEM EDA Forskal 
THEM EDA FRONDOSA (R. Br.) comb. nov. 

Anthistiria frondosa R. Br. Prodr. (1810) 200. 

Themeda arguens Hack, in DC. Monog. Phan. 6 (1889) 657, non 

Stipa arguens Linn. 
Gramen arguens Rumph. Herb. Amb. 6: 15, t. 6, f. 1. 
Amboina, Way tommo, in waste places, Robinson PI. Rumph. Amb. 62, 
August 16, 1913. Celebes, Macassar, Rel. Robins. 2452, July 11, 1913. 

The type of Stipa arguens Linn, was an Indian specimen, 
although the specific name was taken from Rumphius, and 
Gramen arguens Rumph. is cited in the original description. 
The actual type, in the Linnean herbarium, is the form described 
by Hackel as Themeda ciliata (Linn, f.) Hack., to which the 
name Themeda arguens must now be applied. The Linnean 
description was manifestly based on the specimen before him, not 
on the Rumphian illustration ; and, accordingly, the name should 
go with the plant he described. The error in referring the 
Rumphian illustration to the Indian species was a very natural 
one. 

THEMEDA GIGANTEA (Cav.) Hack, in DC. Monog. Phan. 6 (1889) 670. 
Anthistiria gigantea Cav. Ic. 5 (1799) 36, t. 458. 
Calamagrostis Rumph. Herb. 6: 16, t. 6, f. 2. 

This species is not represented in our Amboina collections, but 



90 



RUMPHIUS'S HERBARIUM AMBOINENSE 



the above disposition of Rumphius's Calamagrostis is suggested 
as its probable true position. Linnaeus, Sp. PL ed. 2 (1762) 65, 
referred it to his Schoenus lithospermus=Scleria lithosperma 
Sw., in which he was followed by Burman f., Fl. Ind. (1768) 
19. Willdenow, however, Sp. PL 4 (1805) 315, referred it to 
Scleria tessellata Willd., in which he has been followed by several 
other authors. The plant, as described by Rumphius, has 
nothing to do with Scleria, but is manifestly a coarse grass, and 
it is certainly Anthistiria gigantea Hack., as suggested above. 
The figure is very poor. 

DIGITARI A Scopoli 

DIGITARIA SANGUINALIS (Linn.) Scop. Fl. Cam. ed. 2, 1 (1772) 

52, var. 

Panicum sanguinale Linn. Sp. PI. (1753) 57. 
Gramen caninum Rumph. Herb. Amb. 6: 11. 
Gramen supplex Rumph. 1. c. 12? 
Amboina, in sago swamp near the town of Amboina, Robinson PL Rumph. 
Amb. 50, August 20, 1913. 

This is apparently Gramen caninum Rumph., described as 
having two spikes. Gramen supplex Rumph. is described as 
having three or four spikes. Both appear to be merely forms 
of the polymorphous Digitaria sanguinalis (Linn.) Scop. 

PANICUM Linnaeus 

PANICUM REPTANS Linn. Syst. ed. 10 (1759) 870. 
Panicum prostratum Lam. 111. 1 (1791) 171. 
Gramen anatum Rumph. Herb. Amb. 6: 13. 
Boeton, Rel. Robins. 2496, July 13, 1913; not represented in the Am- 
boina collection. 

In this reduction of Gramen anatum I follow Hasskarl's sug- 
gestion, who refers it with doubt to Panicum prostratum Lam. I 
can see no reason for considering the Rumphian plant other 
than this species ; the Linnean name is, however, the older. 

PANICUM STAGNINUM Retz. Obs. 4 (1786) 17. 

Champeu s. Campee, Rumph. Herb. Amb. 6: 11. 

The description is very brief, but probably Panicum stagni- 
num Retz. is the form intended. Hasskarl has suggested Pani- 
cum limnaeum Steud., but this is very improbable, P. limnaeum 
Steud. being reduced in Index Kewensis to Panicum molle Sw. 
The material considered by Rumphius was from Batavia, Java. 
Koorders, Exkurs. Fl. Java 1 (1911) 129, gives tjampea as the 
Sundanese name for Panicum stagninum Retz. 



GRAMINEAE 



91 



PANICUM PALM AEFOLI U M Koenig in Naturf. 23 (1788) 208. 
Angraecum terrestre alterum Rumph. Herb. Amb. 6: 115? 

The description is rather indefinite, but apparently applies 
to Panicum palmae folium Koenig. The reduction follows Hass- 
karl's suggestion, Neue Schlussel (1866) 171, who thought 
that it was either Panicum palmae folium Koenig or P. nepalense 
Spreng. 

OPLISMENUS Beauvois 

OPLISMENUS COMPOSITUS (Linn.) Beauv. Agrost. (1812) 54. 
Panicum compositum Linn. Sp. PL (1753) 57. 

Hippogrostis amboinica II minor Rumph. Herb. Amb. 6: 14, t. 5, f. 3. 

Linnaeus, Amoen. Acad. 4 (1759) 133, reduced Hippogrostis 
amboinica to Panicum colonum, merely citing the number of 
the plate. There are three figures on the plate : one, Andropogon 
aciculatus Retz., that manifestly Linnaeus did not intend to refer 
to Panicum colonum; two, what I take to be Ischaemum timo- 
rense Kunth, and which does not resemble Panicum colonum; 
and three, what I take to represent Oplismenus compositus 
Beauv., probably the figure that Linnaeus intended to refer to 
Panicum colonum. However, it certainly is not this species. 
Later, Linnaeus did refer t. 6, f. 3, to Panicum patens, Mant. 2 
(1771) 323, apparently following Burman f., Fl. Ind. (1768) 
26, but Panicum patens Burm. f . has nothing to do with Panicum 
patens Linn. Lamarck, Encycl. 4 (1798) 742, refers it to his 
Panicum bromoides, which is presumably the same as Oplis- 
menus burmannii Beauv. Oplismenus compositus Beauv. does 
not appear in our Amboina collections, but 0. burmannii Beauv. 
is represented by Rel. Robins. 16J>5, collected in forests at Ayer 
putri, July 29, 1913. It does not however, agree with 
Rumphius's description of Hippogrostis amboinica or with the 
figure cited above. 

SETARIA Beauvois 

SETARIA FLAVA (Nees) Kunth Rev. Gram. 1 (1829) 46. 
Panicum flavum Nees in Mart. Fl. Bras. 2 (1829) 180. 
- Panicum polystachion Linn. Syst. ed. 10 (1759) 870, Sp. PL ed. 2 
(1762) 82 (type!), non Setaria polystachya Schrad., nec Scheele. 
Panicum vulpinum Linn. Amoen. Acad. 4 (1759) 134 (type!), non 

Willd., nec Setaria vulpina Beauv. 
Gramen vulpinum Rumph. Herb. Amb. 6: 18, t. 7, f. 2B. 
Amboina, Soja road, Lateri, and Gelala, Robinson PI. Rumph. Amb. 41, 
August, September, 1913. 

The Rumphian reference is the whole basis of both Panicum 
vulpinum Linn., which does not appear in Index Kewensis, and 



92 



RUMPH IUS'S HERBARIUM AMBOINENSE 



Panicum polystachion Linn., the former based on "Gramen cari- 
cosum vulpinum" the latter on Vol. VI "t. 7, f. 2B," and er- 
roneously reduced in Index Kewensis to Andropogon caricosum 
L. Both Linnean specific names are invalid in Setaria, although 
both are much older than Panicum flavum Nees. By Burman 
f., Fl. Ind. (1768) 24, it was considered under Panicum poly- 
stachion Linn., and it was also so considered by Loureiro, Fl. 
Cochinch. (1790) 46. The species is very generally considered 
to be a synonym of Setaria glauca (Linn.) Beauv., but is ap- 
parently distinct. 

SETARIA ITALICA (Linn.) Beauv. Agrost. (1812) 51. 
Panicum italicum Linn. Sp. PI. (1753) 56. 

Panicum indicum Rumph. Herb. Amb. 5: 202, t. 75 bis, f. 2. 

The figure is a good representation of one of the commonly 
cultivated forms of this species, with the addition of the panicle 
of another form. It was reduced to Panicum italicum by Lin- 
naeus in Stickman Herb. Amb. (1754) 20, repeated in Amoen. 
Acad. 4 (1759) 130, Syst. ed. 10 (1759) 870, Sp. PL ed. 2 (1762) 
83; followed by Burman f., Fl. Ind. (1768) 24, Loureiro, FL 
Cochinch. (1790) 46, Willdenow, Sp. PI. 1 (1797) 336, and other 
authors. For a critical botanical consideration of the numerous 
varieties and forms of Setaria italica see Hubbard in Am. Journ. 
Bot. 2 (1915) 169-198. 

THUAREA Persoon 

THUAREA INVOLUTA (Forst.) R. Br. ex Steud. Nomencl. ed. 2, 2 (1841) 

682. 

Ischaemum involutum Forst. Prodr. (1786) 73. 
Thuarea sarmentosa Pers. Syn. 1 (1805) 110. 
Gramen roris (litoreum) Rumph. Herb. Amb. 6: 13. 
Amboina, Hatiwe, along the strand, Robinson PL Rumph. Amb. 51, 
September 4, 1913. 

The description is very short, and the identification has been 
made chiefly from the indicated habitat. Hasskarl refers here 
Gramen roris as described on page 12 and is perhaps correct 
in doing so. R. Brown, Prodr. (1810) 197, does not publish 
Thuarea involuta as currently indicated in botanical literature, 
but merely indicates that Ischaemum involutum Forst. pertains 
to the genus Thuarea. 

SPINIFEX Linnaeus 

SPINIFEX LITTOREUS (Burm. f.) Merr. in Philip. Journ. Sci. 7 (1912) 
Bot. 229. 

Stipa littorea Burm. f. Fl. Ind. (1768) 29. 
Stipa spinifex Linn. Mant. 1 (1767) 84. 



GRAMINEAE 



93 



Spinifex squarrosus Linn. Mant. 2 (1771) 300. 
Cyperus littoreus Rumph. Herb. Amb. 6:6, t. 2, f. 2. 

This species is not represented in our Amboina collection. It 
is a very widely distributed grass, occurring on sandy beaches 
in the Indo-Malayan region, and Rumphius gives a good figure 
of the pistillate plant. Cyperus littoreus is, by citation, the type 
of Stipa littorea Burm. f., and the specific name is taken from 
Rumphius; however, it is perfectly evident that Burman f. also 
had actual specimens of the same plant. It is also cited by 
Linnaeus in the original publication of Stipa spinifex Linn., Mant. 
1 (1767) 34, and under Spinifex squarrosus Linn, by Willdenow, 
Sp. PI. 4 2 (1805) 1129; by Loureiro, Fl. Cochinch. (1790) 647; 
and by other authors. 

ORYZA Linnaeus 

ORYZA SATIVA Linn. Sp. PI. (1753) 333. 

Oryza vulgaris Rumph. Herb. Amb. 5: 196. 
Oryza glutinosa Rumph. op. cit. 201. 

Rice is not represented in our Amboina collections. Six forms 
are described under Oryza vulgaris, and three under Oryza 
glutinosa, but all are manifestly cultural forms of the polymor- 
phous species Oryza sativa Linn. Oryza communis sima Lour., 
O. praecox Lour., O. montana Lour., and O. glutinosa Lour., Fl. 
Cochinch. (1790) 215, while based primarily on specimens from 
China or Cochin-China, all have references to Rumphius. 
Linnaeus, by error, refers t. 7k to Oryza, in Amoen. Acad. 4 
(1759) 130. Table 7k is Saccharum officinarum, while table 
75 bis, the plate apparently intended by Linnaeus, represents a 
form of Andropogon sorghum Linn, and Setaria italica Kunth; 
table 75 represents Coix lachryma jobi Linn, and apparently a 
form of Saccharum officinarum, probably the species proposed by 
Hasskarl as Saccharum edule Hassk. The common rice plant is 
not figured by Rumphius. 

CYNODON * Persoon 
CYNODON DACTYLON (Linn.) Pers. Syn. 1 (1804) 85. 
Panicum dactylon Linn. Sp. PI. (1753) 58. 
Capriola dactylon O. Ktze. Rev. Gen. PI. 1 (1891) 764. 
Gramen repens minus Rumph. Herb. Amb. 6: 11. 

This common and widely distributed species is not represented 
in our Amboina collection. Hasskarl has made this reduction of 
Gramen repens minus with doubt, but it is perfectly evident that 
Cynodon dactylon Pers. is the plant described by Rumphius. 



* Retained name, Vienna Code; Capriola Adans. (1763) is older. 



94 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ELEUSIN E Gaertner 

ELEUSINE INDICA (Linn.) Gaertn. Fruct. 1 (1781) 8. 

Cynosurus indicus Linn. Sp. PI. (1753) 72. 

Gramen vaccinum Rumph. Herb. Amb. 6: 9, t. 4, /• 2. 
Amboina, Batoe mera, in ditches and along roadsides, Robinson PI. 
Rumph. Amb. Jf8, July 20 and August 15, 1913. 

This common and widely distributed grass is well represented 
by Rumphius's figure, which was first referred by Linnaeus, Sp. 
PI. ed. 2 (1762) 106, to Cynosurus indicus, in which he has been 
followed by Burman f., Fl. Ind. (1768) 29; by Lamarck, Encycl. 
2 (1786) 187; by Loureiro, Fl. Cochinch. (1790) 59; by Willde- 
now, Sp. PI. 1 (1797) 417 — all under Cynosurus — and by Hass- 
karl, Neue Schliissel (1866) 152, under Eleusine. The figure 
is not cited in modern literature. 

ELEUSINE COROCANA (Linn.) Gaertn. Fruct. 1 (1781) 8, t. 1. 

Cynosurus corocanus Linn. Syst. ed. 10 (1759) 875. 
Goddam Rumph. Herb. Amb. 6: 10. 

Pinicum gramineum Rumph. Herb. Amb. 5: 203, t. 76, f. 2. 

Rumphius figured this species from plants grown in Amboina, 
the seeds having been received by him from Ceylon; it is con- 
sidered by Hooker f . to be a cultivated form of Eleusine indica 
Gaertn. The figure given by Rumphius was first referred to 
this species by Linnaeus, Syst. ed. 10 (1759) 875, repeated in his 
Sp. PI. ed. 2 (1762) 107; followed by Burman f., Fl. Ind. (1768) 
29, by Lamarck, Encycl. 2 (1786) 187, and by Willdenow, Sp. PI. 
1 (1797) 415 — ^all under Cynosurus corocanus Linn. — and finally 
by Hasskarl, Neue Schliissel (1866) 111. Rumphius's figure is 
not generally cited by modern authors. The Javan form very 
briefly described by Rumphius as Goddam can scarcely be other 
than Eleusine corocana Gaertn. 

DACTYLOCTEN I U M Willdenow 

DACTYLOCTEN I U M AEGYPTIUM (Linn.) Richt. PI. Europ. 1 (1889) 68. 

Cynosurus aegyptius Linn. Sp. PI. (1753) 72. 

Gramen vaccinum femina Rumph. Herb. Amb. 6: 9, 10, t. U, f- 1- 

Amboina, Batoe mera, along roadsides, Robinson PI. Rumph. Amb. 4.9, 
August 15, 1912. 

This common grass is fairly well represented by Rumphius's 
figure, although the prolonged rachis is not shown. The habit, 
together with the relatively thick spikes, unquestionably places 
the figure cited with Dactyloctenium rather than with Eleusine 
indica Gaertn. The figure was first referred here by Linnaeus, 
in Stickman Herb. Amb. (1754) 25, Amoen. Acad. 4 (1759) 



GRAMINEAE 



95 



133, Syst. ed. 10 (1759) 875; and later by Loureiro, Fl. Cochinch. 
(1790) 59, by Roxburgh, FL Ind. ed. 2, 1 (1832) 344— all under 
Cynosurus — and finally by Hasskarl, Neue Schliissel (1866) 152, 
as Dactyloctenium aegyptiacum Willd. * The figure is rarely 
cited in modern literature. The species usually appears in 
literature as Dactyloctenium aegyptiacum Willd. 

PHRAGMITES Trinius 

PHRAGM ITES VULGARIS (Lam.) Trin. Fund. Agrost. (1820) 134. 

Arundo vulgaris Lam. Fl. Franc. 3 (1778) 615. 

Arundo phragmites Linn. Sp. PI. (1753) 81. 

Canna palustris Rumph. Herb. Amb. 4: 20, t. 5. 
Amboina, Wakal, near the beach, Robinson PL Rumph. Amb. 44, Novem- 
ber 5, 1913, known as tebu-tebu. 

Rumphius describes Canna palustris as from 12 to 16 feet 
high ; Robinson notes that his plant, cited above, is from 3 to 4 
meters high. Both may be referable to Phragmites karka 
(Retz.) Trin., rather than to P. vulgaris Trin. Loureiro, FL 
Cochinch. (1790) 54, erroneously considers it under Aira arundi- 
nacea Linn. Hasskarl, Neue Schliissel (1866) 71, has reduced 
Canna palustris Rumph., with doubt, to Eulalia japonica Trin. 
=Miscanthus japonicus Anders.; this suggested reduction is 
impossible, however, as the habitat and other data given by 
Rumphius apply to Phragmites, not to Miscanthus. The figure 
of Canna palustris is very crude, and the true position of the 
species has not before been recognized. 

ERAGROSTIS Host 

ERAGROSTIS AM ABILIS (Linn.) W. & A. in Hook. & Arn. Bot. Beech. 
Voy. (1841) 251, excl. descr. 

Poa amabilis Linn. Sp. PI. (1753) 68. 

Poa tenella Linn. Sp. PI. (1753) 69. 

Eragrostis plumosa Link Hort. Berol. 1 (1827) 192. 

Eragrostis tenella R. & S. Syst. 2 (1817) 576. 

Gramen fumi Rumph. Herb. Amb. 6: 11, t. 4, f. 3. 

Bali, Rel. Robins. 2515, 2532, July 7, 1913, but not represented in the 
Amboina collection. 

Gramen fumi was reduced by Linnaeus to Poa tenella Linn,, 
Sp. PL ed. 2 (1762) 101, in which he was followed by Burman 
f., Fl. Ind. (1768) 28, Poiret in Lam ; Encycl. 5 (1804) 85, and 
by various other authors. Roxburgh, Fl. Ind. ed. 2, 1 (1832) 
337, refers it to Poa plumosa Retz.= Eragrostis plumosa Link 
=Eragrostis amabilis (Linn.) W. & A. Poa amabilis Linn, is 
identical with the species commonly known as Eragrostis tenella 



96 



RUMPHIUS'S HERBARIUM AMBOINENSE 



R. & S. and E. plumosa Link. Poa tenella Linn, was based on 
specimens from India and is identical with Poa amabilis Linn. ; 
see Munro in Journ. Linn. Soc. Bot. 6 (1862) 43. It is to be 
noted that in transferring the Linnean specific name to Eragros- 
tis, as E. amabilis, Wight and Arnott describe a form that is 
totally distinct from Poa amabilis Linn, and is Eragrostis unio- 
loides Nees. The specific name amabilis manifestly belongs 
with the plant originally described by Linnaeus, not with Era- 
grostis unioloides Nees. 

BAMBUSEAE 

Under the names Arundo and Arundarbor, Rumphius has 
described a number of species and forms of bamboo, which have 
been very imperfectly understood by later authors, although 
many species have been based, wholly or in part, on Rumphius's 
descriptions. These numerous species, chiefly proposed by 
Loureiro, Roemer and Schultes, and Miquel, have remained 
doubtful, almost without exception, to the present time. The 
Amboinian material presents four distinct species, three of which 
were described by Rumphius, and this material has enabled me 
to solve several problems in synonymy and definitely to 
determine the status of a number of specific names in various 
genera that have been based on Rumphius. The others have 
been interpreted from the data given by Rumphius, but much 
field work is necessary before a sufficient amount of data is 
available properly to interpret the numerous forms named by 
Rumphius. In this connection it is to be noted that it has 
been possible definitely to determine the status of every species 
of bamboo described by Blanco from the Philippines, a total 
of eight, from field work in connection with Blanco's descriptions 
and the native names cited by him; yet Blanco's descriptions, 
on the average, are decidedly inferior to those of Rumphius. 
Up to 1900 not one of Blanco's species had been definitely placed, 
those that were considered at all appeared in literature as 
species of doubtful status. It> is confidently expected that the 
status of most of the species based on Rumphius can be definitely 
settled by following the policy adopted in the Philippines in 
working out the identity of Blanco's species. The specimens 
cited below, with the exception of Bambusa spinosa Roxb., of 
which there were no duplicates, have been critically examined 
by Mr. J. Sykes Gamble, to whom I am under obligations for 
his valuable notes, both in connection with the identity of the 
forms with the Rumphian descriptions and with the names in 
current use for the several species. 



GRAMINEAE 



97 



BAM B USA Schreber 

BAMBUSA SPINOSA Roxb. Hort. Beng. (1814) 25 (type!). 

Bambusa spinosa Blume ex Nees in Flora 8 (1825) 580. 

Bambusa blumeana Schultes f. Syst. 7 2 (1830) 1343. 

Bambusa teba Miq. Fl. Ind. Bat. 3 (1857) 418 (type!). 

Arundarbor spinosa Rumph. Herb. Amb. 4: 14, t. 3. 
Amboina, back of the town of Amboina, Robinson PI. Rumph. Amb. 608, 
July 17, 1913, locally known as bambu duri. 

There is no question that this is Arundarbor spinosa of 
Rumphius, and that it is identical with the widely distributed 
Malayan Bambusa blumeana Schultes f. However, Bambusa 
spinosa Roxb. was based wholly on Rumphius in the original 
place of publication, Hortus Bengalensis (1814) 25, by citation 
of "H. A. 4. t. 3;" that is, Herbarium Amboinense Vol. IV, t. 3. 
Robinson,* however, states (p. 418) : "Bambusa spinosa Roxb., 
as typified by Herb. Amb. 4 : pi. 3, is probably not B. arundinacea 
Willd., and there are various points in Rumphius's description, 
which oppose its identification as B. Blumeana Schultes f." 
Munro t states under Bambusa blumeana Schultes f . : "Blume 
states that this plant is spinous; and there are some traces 
of spines on the specimens I have seen; but there is nothing 
to indicate that it is as spinous as B. agrestis of Poir. and 
Arundarbor spinosa of Rumph. are described to be." From 
an examination of the Amboina specimen, Rumphius's descrip- 
tion, and with a definite knowledge of this plant as it grows 
in the field, for it is one of the commonest bamboos in the 
Philippines, I am thoroughly convinced that Arundarbor spinosa 
is identical with Bambusa blumeana Schultes f., but under our 
rules of nomenclature, the oldest valid specific name is Bambusa 
spinosa Roxb. The plant is remarkable for the dense thicket 
of interlaced, stiff, very spiny branches that surrounds the 
lower parts of the culms, and these are well represented by 
Rumphius's figure. Ordinary herbarium specimens, taken from 
the upper leafy branches often do not present spines, or at 
least only few and greatly reduced ones, which accounts for 
Munro's statement, as he saw only herbarium specimens and 
did not know the plant in the field. 

By Loureiro Arundarbor spinosa was reduced to his Arundo 
agrestis, Fl. Cochinch. (1790) 57, but Arundo agrestis Lour, was 
based on specimens from Cochin-China, and apparently is Bam- 

* Robinson, C. B. Roxburgh's Hortus Bengalensis. Philip. Journ. Sci. 
7 (1912) Bot. 410-419. 

f Trans. Linn. Soc. 26 (1870) 102. 

144971 7 



98 



RUMPHIUS'S HERBARIUM AMBOINENSE 



busa arundinacea Retz., not B. blumeana Schultes f . Loureiro's 
species was transferred to Bambusa as B. agrestis Poir. in Lam. 
Encycl. 8 (1808) 704. Munro includes Loureiro's species in 
the synonymy of Bambusa blumeana Schultes f., with doubt; 
I am convinced that it should be transferred to Bambusa arun- 
dinacea Retz. Bambusa teba Miq. is based wholly on Rumphius, 
for Miquel cites as synonyms Arundo agrestis Lour, and Bam- 
busa agrestis Poir., specifically excluding all data given by 
these authors except the Rumphian synonym, Arundarbor spi- 
nosa Rumph. 

BAMBUSA ATRA Lindl. in Penny Cyclop. 3 (1835) 357 (type!). 
Bambusa prava Lindl. 1. c. (type!). 
Bambusa picta Lindl. 1. c. (type!). 

Leleba alba, nigra, prava, et lineata Rumph. ex R. & S. Syst. 7 2 (1830) 
1345, 1346. 

Bambusa tenuis Munro in Trans. Linn. Soc. 26 (1868) 119 (type!). 
T eleba rumphiana Kurz Cat. Hort. Bogor. (1866) 20 (type!). 
bambusa rumphiana Kurz in Journ. As. Soc. Beng. 39 2 (1870) 86. 
Bambusa lineata Munro in Trans. Linn. Soc. 26 (1868) 118 (type!). 
Arundarbor tenuis Rumph. Herb. Amb. 4: 1, t. 1 (inch alba, nigra, 
lineata, et prava). 
Amboina, Way tommo, Robinson PI. Rumph. Amb. 32, August 18, 1913; 
Caju poeti, PL Rumph. Amb. 33, August 2, 1913; Lateri, Robinson PI. 
Rumph. Amb. 3Jf, September 5, 1913; all known as leleba. 

The synonymy of this species is rather complicated, but Bam- 
busa atra Lindl., 1835, based on Leleba nigra, is apparently 
the oldest valid specific name. It is to be noted that the several 
forms described by Rumphius under the designations Leleba 
nigra, prava, picta, lineata, and amahussana are repeated by 
Roemer and Schultes, Syst. 7 2 (1830) 1345, 1346, under the 
Rumphian names in a note following Bambusa verticillata 
Willd. "Publication" was not intended, nor have Roemer and 
Schultes's names been recognized as published. 

Regarding the actual specimens cited above, Mr. Gamble writes 
as follows: "I make Nos. 32, 33, 34 all Bambusa rumphiana 
Kurz. In my Bambuseae of British India* I described this as 
Bambusa lineata Munro, but after publication Sir D. Brandis 
pointed out to me that Munro's B. lineata only referred to one 
of Rumphius's varieties, while Kurz's name included them all. 
So I agree with him, but it is best to take Kurz's name. No. 
32, I agree, fits best with Rumphius's Arundo arbor tenuis 
nigra, and No. 33 with A. tenuis alba; so too No. 34, with very 
broad leaves, will do for A. tenuis prava." 



* Ann. Bot. Gard. Calcutta 7 (1896) 1-133, t. 1-118. 



GRAMINEAE 



99 



Historically, the first reference to Arundo arbor tenuis is 
found in Loureiro's Flora Cochinchinensis (1790) 58, where he 
refers it to his Arundo multiplex. However, Arundo multiplex 
Lour, was primarily based on specimens from Cochin-China and 
of course must be interpreted from Cochin-China material. It 
is Bambusa multiplex Raeusch, a species of doubtful status, 
which Munro suggests closely approaches Bambusa nutans Wall. 
Willdenow erroneously referred the Rumphian species to his 
Bambusa verticillata, Sp. PL 2 (1799) 245, in which he was 
followed by many authors. Willdenow's species, however, was 
based on actual specimens, has nothing to do with the Rumphian 
plant, and is Gigantochloa verticillata Munro. All of the 
synonyms cited above, except Bambusa rumphiana Kurz and 
B. lineata Munro, have previously been considered to represent 
species of doubtful status. 

Var. AMAHUSSANA (Lindl.) comb. nov. 

Bambusa amahussana Lindl. in Penny Cycl. 3 (1835) 357 (type!). 
Bambusa rumphiana Kurz, var. amahussana Gamble in lit. 
Arundarbor amahussana (i. e. Leleba amahussana) Rumph. Herb. 
Amb. 4: 3. 

Amboina, Amahoesoe, Robinson PI. Rumph. Amb. 35, July 30, 1913, near 
the beach, 2 to 3 m high, often reclining, locally known as bambu nitu. 

Regarding this form Mr. Gamble writes as follows: "No. 35 
is identified, with much probability, with Arundarbor tenuis 
amahussana. It comes from Amahoesoe and differs from the 
others in the small spikelets. It is, I think, quite a distinct 
variety, but only a variety, for dissection of the flowers pre- 
sents no differential characters. It is characterized by its very 
small spikelets." 

In connection with the various forms of this species it is 
well to note that Kurz* has enumerated five varieties under 
the Malayan names, Leleba dyahat, pootee, ietam, tootool, and 
soorat. 

BAMBUSA VULGARIS Schrad. in Wendl. Collect. PI. 2 (1810) 26, t. U7. 
Bambusa fera Miq. PI. Ind. Bat. 3 (1855) 418 (type!). 
Bambusa vasaria Munro in Trans. Linn. Soc. 26 (1870) 122 (type!)? 
Arundarbor fera Rumph. Herb. Amb. 4: 16, t. 
Arundarbor vasaria s. Bulu Java Rumph. Herb. Amb. 4: 8? 

Arundarbor fera Rumph. is the whole basis of Bambusa fera 
Miq., and it seems to be typical Bambusa vulgaris Sclirad., a 
disposition of it suggested by Miquel. Loureiro, FL Cochinch. 
(1790) 57, referred it to Arundo mitis Lour., a species typified 



* Ind. Forester 1 : 341. 



100 



RUMPHIUS'S HERBARIUM AMBOINENSE 



by Cochin-China specimens. Bambusa mitis Poir., in Lam. 
Encycl. 8 (1808) 704, is merely a new name for Arundo mitis 
Lour. Loureiro's specimens were from plants cultivated in 
Cochin-China, and the species may well be the same as Bambusa 
vulgaris Schrad., in which case the specific name mitis will have 
priority. In proposing Bambusa fera, Miquel explicitly excludes 
from the synonyms cited all but the reference to Rumphius. 

Arundarbor vasaria s. Bulu Java Rumph. is the whole basis 
of Bambusa vasaria Munro and is probably the same as Bam- 
busa vulgaris Schrad. Munro suggested that it might be the 
same as Bambusa balcooa Roxb., but the distribution of the 
latter, known only from India, makes this reduction an impos- 
sible one. Schultes quotes it under Bambusa arundinacea Retz., 
Loureiro under Arundo bambos Lour., while Miquel, Fl. Ind. 
Bat. 3 (1855) 417, thinks that it may be Bambusa vulgaris 
Schrad. Arundarbor vasaria cho Rumph. 4: 10, may also belong 
here. Under Arundarbor vasaria Rumphius describes several 
other forms, on which, fortunately, subsequent botanists have 
based no binomials, either wholly or in part. From the data at 
present available it is impossible to determine their status, and 
they are accordingly not listed here. The full list is given by 
Hasskarl, Neue Schlussel (1866) 69-71. 

BAM E USA VULGARIS Schrad. var. STRIATA (Lodd.) Gamble in Ann. 
Bot. Gard. Calcutta 7 (1896) 44. 
Bambusa striata Lodd. ex Lindl. Penny Cyclop. 3 (1835) 357. 
Arundarbor fera elegantissima Rumph. Herb. Amb. 4: 16. 

The description given by Rumphius applies unmistakably to 
this form of Bambusa vulgaris Schrad., which is widely culti- 
vated in the tropics of the Old World for ornamental purposes. 

BAMBUSA EXCELSA Miq. Fl. Ind. Bat. 3 (1855) 418 (type!). 
Arundarbor maxima Rumph. Herb. Amb. 4: 12. 

Loureiro, Fl. Cochinch. (1790) 58, referred Arundarbor maxi- 
ma Rumph. to Arundo maxima Lour., but his description was 
based on Cochin-China specimens, which probably represent a 
species quite different from the one Rumphius described. Poiret, 
in Lam. Encycl. 8 (1808) 704, transferred it to Bambusa maxima 
Poir. Miquel, however, in proposing the name Bambusa ex- 
celsa, specifically excludes from the descriptions of Loureiro 
and Poiret everything except the references to Rumphius; Me- 
locanna excelsa Roep., in Trin. Clav. Agrost. (1822) 105, which 
probably goes with Arundarbor maxima Lour., is cited as a 
doubtful synonym. Munro has suggested that Bambusa excelsa 



GRAMINEAE 



101 



Miq. is a synonym of Gigantochloa verticillata Munro. The 
native names given by Rumphius are bulu sammet, bulu gantang, 
bulu wani besaar, terin maysele, and tabatiko sammat. I suspect 
that it is, at least in part, Dendro calamus giganteus Munro. 

GIGANTOCHLOA Kurz 

GIGANTOCHLOA ASPERA (Schultes) Kurz ex Koord. Exkurs. Fl. Java 
1 (1911) 176. 

Bambusa aspera Schultes Syst. 7 2 (1830) 1352 (type!). 
Arundarbor aspera Rumph. Herb. Amb. 4: 11, t. 2. 

Munro,* sub Gigantochloa atter Kurz, states: "Kurz, in his 
notes, identifies this species [Gigantochloa atter] with Bambusa 
aspera and B. bitung Roem. & Sch." to which Kurz f rejoins : 
"I do not understand this interpretation in which I am said to 
have identified 2 such species, as those alluded to, which differ 
toto coelo! As far as I am aware I have identified B. aspera 
with B. Bitung, but surely not these two with B. atter. The one 
is (sententia Munroana) a Dendro calamus, the other a Gigan- 
tochloa." He then adds a diagnosis of Bambusa aspera as he 
understands the species, but which is not published under Gigan- 
tochloa. The range is given as "Indian Archipelago, from the 
Moluccas to Singapore." Bambusa aspera R. & S. is not men- 
tioned by Gamble,} but Bambusa bitung R. & S. is placed as a 
probable synonym under Dendrocalamus flagellifer Munro. 
Camus § likewise does not mention Bambusa aspera R. & S., 
but follows Munro and Gamble in the disposition of Bambusa 
bitung R. & S. I have found no transfer of the species to 
Gigantochloa antedating that made by Koorders, cited above. 
The species, which is one of doubtful status, may be a 
Dendrocalamus. 

DINOCHLOA Biise 

DINOCHLOA sp.? 

Boeloe rottang Rumph. Herb. Amb. 5: 119. 

The description manifestly applies to some scandent species 
of bamboo, which might be either of the genus Dinochloa or 
Schizostachyum. Hasskarl, Neue Schlussel (1866) 102, referred 
it with doubt to Dinochloa tjankorreh Buse=Z). scandens 0. 
Ktze., but in the absence of material from Amboina representing 
any scandent bamboo, all reductions of Boeloe rottang must be 
uncertain. 

* Trans. Linn. Soc. 26 (1870) 125. 

fJourn. As. Soc. Beng. 39 2 (1870) 87. 

$Ann. Bot. Gard. Calcutta 7 (1896) 1-133, t. 1-118. 

§ Les Bambusees (1913) 1-215. 



102 



RUMPHIUS'S HERBARIUM AMBOINENSE 



SCHIZOSTACHYUM Nees 

SCHIZOSTACHYUM BRACH YCLADU M Kurz in Journ. As. Soc. Beng. 
39 3 (1870) 89. 
Arundarbor cratium Rumph. Herb. Amb. 4: 5. 
Amboina, Batoe merah, Robinson PI. Rumph. Amb. 36, August 11, 1913; 
Lateri, Robinson PI. Rumph. Amb. 37, August 26, 1913, locally known as 
bulu jawa and bulu seru, in dense clumps, culms from 5 to 10 meters high, 
the internodes 35 to 90 cm in length. 

In regard to the numbers cited, Mr. Gamble writes as fol- 
lows: "I think that you have correctly identified these with 
Arundarbor cratium. The specimens are quite good ones of 
Schizostachyum brachycladum Kurz, and agree with a series 
of specimens from Java and the Philippines as well as with 
specimens from cultivated plants in the Calcutta garden." 
Loureiro, Fl. Cochinch. (1790) 58, referred Arundarbor cratium 
to Arundo fax Lour., a species based on Cochin-China specimens 
and probably not the same as the Amboina form. Melocanna 
humilis Roep., ex Trin Clav. Agrost. (1822) 105, is merely 
a new name for Arundo fax Lour. Beesha humilis Kunth, 
Enum. 1 (1833) 434, and Beesha fax R. & S., Syst. 7 2 (1830) 
1336, are also synonyms of Arundo fax Lour, and like Loureiro's 
species are not to be interpreted by the latter's erroneous reduc- 
tion of Arundarbor cratium Rumph. Rumphius briefly describes 
six forms following Arundarbor cratium, but the descriptions 
are not sufficiently definite to warrant the reduction of any 
of them. They may or may not be variants of Schizostachyum 
brachycladum Kurz. See also under Schizostachyum sp. 
(Bambusa longinodis Miq.). 

SCHIZOSTACHYUM sp. 

Bambusa longinodis Miq. Fl. Ind. Bat. 3 (1855) 418 (type!). 
Arundarbor spiculorum Rumph. Herb. Amb. 4: 7. 

Miquel's species was based wholly on Rumphius, as he speci- 
fically includes from Bambusa tabacaria Poir. {Arundo taba- 
caria Lour.) only the reference to Rumphius. Loureiro, Fl. 
Cochinch. (1790) 68, quotes Rumphius's species as a synonym 
of Arundo tabacaria Lour., a species based on Cochin-China 
specimens undoubtedly different from the Amboina plant; Bam- 
busa tabacaria Poir. is merely a new name for Loureiro's species 
and must be interpreted from Loureiro's description, not from 
Rumphius. Bambusa longinodis Miq. is unquestionably a 
species of Schizostachyum closely allied to the Philippine 
Schizostachyum lima (Blanco) Merr. (S. hallieri Gamble) and 



CYPERACEAE 



103 



to Schizostachyum brachycladum Kurz. It is barely possible 
that the Lateri specimen, cited above under Schizostachyum 
brachycladum Kurz (PL Rumph, Amb. 37) and described by 
Doctor Robinson as having internodes 60 to 90 cm in length, 
may be Arundarbor spiculorum Rumph. In PL Rumph. Amb. 
36, also cited under Schizostachyum brachycladum Kurz, the 
internodes are from 35 to 40 cm in length, but the branches, 
leaves, inflorescence, and spikelets are identical in both 
specimens. 

BAMBUSEAE OF UNCERTAIN STATUS 

The following forms of bamboo described by Rumphius are undetermin- 
able, and fortunately, with possibly one exception, they have not been made 
the basis of binomials by subsequent authors: 

Arundarbor fera silvestris Rumph. Herb. Amb. 4: 16. 

This is probably the same as Bambusa spinosa Roxb. 
Arundarbor ferae adf. Rumph. Herb. Amb. 4: 18. 
Arundarbor fera nigra Rumph. Herb. Amb. 4: 18. 

This may be Bambusa nigra Lodd., where it has been placed by several 
authors. 

Arundarbor fera s. cha Rumph. Herb. Amb. 4: 18. 
Arundarbor fera 6. bulu tsjatjar Rumph. Herb. Amb. 4: 18. 
Arundarbor fera s. Arundo japanica Rumph. Herb. Amb. 4: 18. 

CYPERACEAE 

KYLLINGA Rottboell 
KYLLINGA MONOCEPHALA Rottb. Descr. et Ic. PI. (1773) 13, t. U, /• U. 
Gramen capitatum Rumph. Herb. Amb. 6:8, t. 3, f. 2. 
Amboina, near the town of Amboina, Robinson PL Rumph. Amb. £29, 
July 25, 1913, in meadows near sea level, associated with Kyllinga brevi- 
folia Rottb. (Rel. Robins. 1897). 

The figure presents a good habit sketch of what might be 
either Kyllinga monocephala Rottb. or K. brevifolia Rottb., but 
as the inflorescence is described as white, I have referred the 
Rumphian figure and description to Kyllinga monocephala; the 
inflorescence of K. brevifolia Rottb. is usually, if not always, 
green instead of white. Historically, the Rumphian figure was 
first considered by Linnaeus, in Stickman Herb. Amb. (1754) 
25, Amoen. Acad. 4 (1759) 133, where it is erroneously reduced 
to Scirpus glomeratus Linn. Burman f., Fl. Ind. (1768) 18, 
and Loureiro, Fl. Cochinch. (1790) 41, erroneously refer it to 
Schoenus coloratus Linn. Lamark, however, Encycl. 3 (1789) 
366, correctly reduced it to Kyllinga monocephala Rottb., this 
reduction being very generally accepted by subsequent authors. 



104 



RUMPHIUS'S HERBARIUM AMBOINENSE 



REM IREA Aublet 

REM I REA MARITIMA Aubl. PI. Guin. (1775) 45, t. 16. 
Cyperus longus Rumph. Herb. Amb. 6: 5, t. 2, f. 1. 
Amboina, Hatiwe, Robinson PL Rumph. Amb. 436, September 4, 1913, 
growing on the strand. 

The specimen agrees entirely with Rumphius's description and 
with the excellent figure given by him. The identity of Cyperus 
longus with Remirea maritime has not been previously sug- 
gested. By Kunth, Enum. 2 (1837) 94, it was erroneously 
reduced to Cyperus kyllingioides Vahl, following Roemer and 
Schultes, Mant. 2 (1824) 98, who, however, cite by error Rheede's 
Hortus Malabaricus, rather than Rumphius. This reduction is 
followed by Hasskarl, Neue Schlussel (1866) 152. 

PYCREUS Beauvois 

PYCREUS O DO RAT US (Linn.) Urb. Symb. Antil. 2 (1900) 164. 
Cyperuu odoratus Linn. Sp. PI. (1753) 46, excl. syn. Sloane. 
Pycreus polystachyus Beauv. Fl. Oware et Benin 2 (1807) 48, t. 86, 
f. 2. 

Cyperus polystachyus R. Br. Prodr. (1810) 214. 
Cyperus floridus II mas Rumph. Herb. Amb. 6: 2, t. 1, f. 2. 
Amboina, near the town of Amboina, along sandy beaches, Robinson PL 
Rumph. Amb. 431, August 22, 1913. 

Hasskarl, Neue Schlussel (1866) 151, refers both figures 1 
and 2 of table 1 to Cyperus rotundus Linn. Figure 1 is certainly 
correctly referred by him, but figure 2 and the description of 
Cyperus floridus mas appear to me to agree better with the 
common and widely distributed Pycreus odoratus Urb. 

CYPERUS Linnaeus 

CYPERUS ROTUNDUS Linn. Sp. PI. (1753) 45. 

Cyperus rotundus bulbosus sive legitimus Rumph. Herb. Amb. 6: 1, 

t. 1, f. 1. 

This common species is not represented in our Amboina col- 
lections, but there is not the slightest doubt as to the correctness 
of this reduction, which follows both Kunth and Hasskarl. 

ELEOCH ARIS R. Brown 

ELEOCH ARIS DULCIS (Burm. f.) Trin. ex Henschel Vita Rumph. (1833) 
186. 

Andropogon dulcis Burm. f. Fl. Ind. (1768) 219 (type!), 
Hippuris indica Lour. Fl. Cochinch. (1790) 16. 
Carex tuberosa Blanco Fl. Filip. (1837) 35. 
Cyperus dulcis Rumph. Herb. Amb. 6: 7, t. 3, f. 1. 



CYPERACEAE 



105 



This species is not represented in our Amboina collections, 
but the Rumphian figure unmistakably represents a plant allied 
to Eleocharis tuberosa Schultes, which is possibly but a tuber- 
bearing form of Eleocharis plantaginoidea (Rottb.) W. F. Wight 
(Scirpus plantaginoides Rottb., Eleocharis plantaginea R. Br., 
Scirpus plantagineus Retz.). Rumphius's Cyperus dulcis is in 
part the basis of Burman's Andropogon dulcis, this author also 
citing a figure in Plukenet. There is no indication that Burman 
had actual specimens, the Rumphian reference is the first one 
given, and the specific name is taken from Rumphius ; therefore, 
I have interpreted Cyperus dulcis as the type. In Index 
Kewensis Andropogon dulcis Burm. f. is reduced to Sorghum 
vulgare. The Malayan form very closely approaches the one 
cultivated in southern China, known in Canton as maa tai, that 
is, typical Eleocharis tuberosa (Roxb.) Schultes. Loureiro's 
description of Hippuris indica apparently applies to the wild 
form of this species, as he describes the tubers as small and 
pilose; those of maa tai are smooth and from 2.5 to 4.5 cm in 
diameter. Loureiro quotes Cyperus dulcis Rumph. as a synonym 
of Hippuris indica Lour. It is to be noted that Eleocharis dulcis 
Trin. does not appear in Index Kewensis. 

FIMBRISTYLIS Vahl 

FIM BR I STY LIS SET ACE A Benth. in Hook. Lond. Journ. Bot. 2 (1843) 
239. 

Gramen polytrichum Rumph. Herb. Amb. 6: 17, t. 7, f. 1. 
Amboina, Koesoekoesoe sereh, Robinson PI. Rumph. Amb. b38, August 
12, 1913, along roadsides, altitude about 250 meters. 

The figure given by Rumphius presents a good habit sketch 
that might with almost equal propriety be referred to Fimbri- 
stylis setacea Benth., to F. polytrichoides R. Br., to F. acuminata 
Vahl, or to any similar tufted species with slender leafless 
stems and solitary terminal spikelets. As Fimbristylis setacea 
Benth. is the only species of this type represented in the Amboina 
collection, I have interpreted it as Gramen polytrichum Rumph. 
Linnaeus, through error, reduced it to Eriocaulon setaceum 
Linn., Amoen. Acad. 4 (1759) 134, Syst. ed. 10 (1759) 880, 
followed by Loureiro, FT. Cochinch. (1790) 60. Willdenow, 
Sp. PL 1 (1797) 295, referred it to Scirpus polytrichoides 
Retz. = Fimbristylis polytrichoides R. Br., which disposition of it 
has been followed by various authors. There is nothing in 
the Rumphian description to indicate to which of the species 
of Fimbristylis discussed above it can be referred; from the 



106 



RUMPHIUS'S HERBARIUM AMBOINENSE 



habitat cited by Rumphius, the indications are that Fimbristylis 
setacea Benth., rather than F. polytrichoides R. Br., is the 
correct disposition of it; it is certainly not Fimbristylis acumi- 
nata Vahl. 

GAHNIA Forster 

GAHNIA RAWACENSIS (Kunth) Steud. Syn. PL Cyp. (1855) 164? 
Lamprocarya rawacensis Kunth Enum. 2 (1837) 333. 
Carex culmaris Rumph. Herb. Amb. 6: 21? 

Amboina, Amahoesoesoe, Robinson PL Rumph. Amb. September 18, 
1913, on hillsides, altitude about 150 meters. 

The identity of the specimen with either Gahnia rawacensis 
Steud. or Carex culmaris Rumph. is very doubtful. The speci- 
mens somewhat resemble Gahnia tristis Nees, but are very much 
smaller and differ in various details. They do not, however, 
agree entirely with the rather imperfect description of Gahnia 
rawacensis Steud., the type of which was from Rawak Island, 
in the Moluccas; it has been reduced, apparently erroneously, 
to Gahnia aspera Spreng. Hasskarl, Neue Schliissel (1866) 
155, refers the Rumphian plant to Gahnia javanica Z. & M. 
Doctor Robinson, who suggested the identity of the plant with 
Carex culmaris, states : " Identification considered very doubtful, 
but so must any identification be." 

HYPOLYTRUM Richard 

HYPOLYTRUM LATI FOLIUM Rich, ex Pers. Syn. 1 (1805) 70. 
Carex laevis minor Rumph. Herb. Amb. 6: 21. 
Amboina, Batoe gadjah, Robinson PI. Rumph. Amb. £32, August 5, 1913, 
near streams, altitude about 150 meters. 

This reduction of Carex laevis minor to Hypolytrum latifolium 
Rich, has not prevously been suggested. Rumphius's descrip- 
tion applies closely to the species as here interpreted. Hasskarl, 
Neue Schliissel (1866) 155, has suggested that it might be a 
form of Cyperus polystachyus Rottb. =Pycreus odoratus Urb., 
but that reduction is an impossible one. 

SCIRPIODENDRON Zippel 

SCIRPIODENDRON GHAERI (Gaertn.) Merr. in Philip. Journ. Sci. 9 
(1914) Bot. 268. 

Chionanthus ghaeri Gaertn. Fruct. 1 (1781) 190, t. 29, f. a-e; Boerl. 

in Journ. Linn. Soc. Bot. 31 (1896) 246. 
Scirpiodendron costatum Kurz in Journ. As. Soc. Beng. 38 2 (1869) 85. 
Scirpiodendron pandani forme Zipp. ex Kurz L c. 

Scirpiodendron sulcatum Miq. 111. PL Archipel. Ind. (1871) 65, t. 28. 



CYPERACEAE 



107 



Pandanus caricosus Spreng. Syst. 3 (1826) 897 (type!); Kunth 
Enum. 3 (1841) 98, non Pandanus caricosus Kurz in Journ. Bot. 
5 (1867) 100, t. 62, f. 1-3, nec Warb. in Engl. Pflanzenreich 3 
(1900) 83. 

Pandanus caricosus Rumph. Herb. Amb. 4: 154. 
Amboina, Paso, Robinson PI. Rumph. Amb. 437, November 25, 1913, along 
margins of streams near the beach. 

Pandanus caricosus Rumph. is unquestionably a Scirpioden- 
dron, although the leaves, as described, are unusually long for 
S. ghaeri Merr. The description of the infructescence and 
fruit, however, is unquestionably Scirpiodendron: "Ejus fructus 
raro in conspectum venit, atque in peculiari progerminat petiolo, 
pedem circiter alto, & supra radicem elevato, qui foliis cingitur 
tribus, in triangulo positis, uti in Cypero." Pandanus caricosus 
Spreng. is based absolutely on the Rumphian description, and 
the species as described by him, after Rumphius, is repeated 
by various other authors. Kurz,* however, described a true 
Pandanus, under the name Pandanus caricosus, which was ac- 
cepted by Warburg f as a valid species. As Kurz's name is 
invalidated by Pandanus caricosus Spreng., I propose the new 
specific designation Pandanus kurzii, for Pandanus caricosus Kurz, 
non Spreng. 

MAPANIA Aublet 

MAPAN1A MACROCEPHALA (Gaudich.) K. Sch. ex Warb. in Engl. Bot. 
Jahrb. 13 (1891) 265. 
Hypolytrum macrocephalum Gaudich. in Freyc. Voy. Bot. (1826) 414. 
Lepironia macrocephala Miq. 111. Fl. Archipel. Ind. (1871) 64, t. 27. 
Carex laevis major Rumph. Herb. Amb. 6: 21. 
Amboina, Batoe gadjah, Robinson PI. Rumph. Amb. 4.30, August 5, 1913, 
along streams, altitude about 150 meters. 

Hasskarl, Neue Schliissel (1866) 155, has suggested that 
Carex laevis major of Rumphius is Pandanophyllum palustre 
Hassk., but Rumphius's description does not agree, especially 
in the statement as to the stalk of the inflorescence bearing 
two or three long leaves. It is, with very little doubt, Mapania 
macrocephala K. Sch. 

SCLERIA Bergius 

SCLERIA LITHOSPERMA Sw. Prodr. (1788) 18. 

Carex amboinica II minor Rumph. Herb. Amb. 6: 20. 
Amboina, Wakeroe, Robinson PI. Rumph. Amb. 435, near the seashore, 
October 17, 1913. 

* Journ. Bot. 5 (1867) 100, t. 62, f. 1-3. 
fEngl. Pflanzenreich 3 (1900) 83. 



108 



RUMPHIUS'S HERBARIUM AMBOINENSE 



The identity of Carex amboinica minor Rumph. with Scleria 
lithosperma Sw. is very probable. Hasskarl has suggested, 
Neue Schliissel (1866) 155, that it is Scleria trialata Poir., a 
species originally described from Madagascar. The specimen 
from the Moluccas, determined by Brongniart, in Duperry's Voy. 
Bot. (1829) 165, and referred by Kunth with doubt to Scleria 
trialata Poir., is probably Scleria bancana Miq. 

SCLERIA BANCANA Miq. Fl. Ind. Bat. Suppl. (1861) 602. 

Carex amboinica I major Rumph. Herb. Amb. 6: 20, t. 8, f. 1? 
Amboina, Gelala and Soja road, Robinson PL Rumph. Amb. Jf33, August, 
September, 1913, on barren, grassy hillsides. 

The identification follows a suggestion made by Doctor Robin- 
son, who considers this plant as probably representing the 
Rumphian species. It is to be noted, however, that Rumphius's 
figure and description of Carex amboinica major call for a plant 
with pseudoverticillate leaves, while in Scleria bancana Miq. the 
leaves are all alternate and distant. It is suspected that the 
plant Rumphius intended is Scleria sumatrensis Retz., S. scro- 
biculata Nees, or some very closely allied form, such as S. 
multifoliata Boeckl. Historically the first reference to Rumphius 
is that by Linnaeus, in Stickman Herb. Amb. (1754) 25, 
where under t. 8 he includes, by error, with Carex amboinica, 
Lithospermum amboinicum and Arundinella; for the latter two 
t. 9 was intended, and the same error is repeated in Amoen. 
Acad. 4 (1759) 134. Burman f., Fl. Ind. (1768) 19, based his 
Schoenus paniculatus partly on Carex amboinica Rumph., but 
the type was a specimen from Java, and the species must be 
interpreted from the Javan specimen. It is supposed to include 
both Scleria sumatrensis Retz. and S. alata Thw. The Rumphian 
figure and description have been referred by other authors to 
Scleria flagellum Sw., S. tessellata Brongn., and S. approximata 
Hassk. 

CYPERACEAE of uncertain status 

Cyperus rotundus Rumph. Herb. Amb. 6: 4 quoad humilior, altior, gramen 
bufonum, et III inodorus. 

The descriptions of these four forms manifestly pertain to the Cyperaceae, 
but the data given are too vague to warrant the definite reduction of them 
to any particular genus or species. For the most part, apparently, species of 
Cyperus and Fimbristylis are intended. 

Carex amboinica III Rumph. Herb. Amb. 6: 20. 

A coarse sedge from Bali, there known as tallan tallan; probably one of 
the large species of Cyperus. Hasskarl, Neue Schliissel (1866) 155, thought 
that it might be Rhynchospora aurea Vahl=i?. corymbosa (Linn.) Britt. 



PALMAE 



109 



PALMAE 

The Amboina collections made by Doctor Robinson have thrown 
comparatively little light on the identity of the numerous species 
of palms described by Rumphius, for the reason that very few 
palms were secured by him. Probably in no other group of 
plants is there more confusion among those species based by 
later authors on Rumphius or more species of uncertain status 
than in this family. In the genus Calamus numerous species 
that were based wholly on the Rumphian descriptions and figures 
are quite unrecognizable; their positions within the genus and 
their relationships with other forms are quite undeterminable. 
In the genus Metroxylon it is impossible to determine, from 
material at present available, whether one somewhat polymor- 
phous species or several closely allied ones are represented. In 
order definitely to settle many cases of uncertain nomenclature 
and to determine the true characters of many species that have 
been based wholly on Rumphius, a much more extensive 
botanical exploration of the Moluccas, especially of Amboina, is 
necessary, and in no group of plants is this more important 
than in the Palmae. 

Beccari has recently given us critical and beautifully illus- 
trated monographs of the genera Calamus and Daemonorops* 
yet of the fifteen species belonging in these two genera, described 
and for the most part figured by Rumphius, he was able de- 
finitely to recognize and to connect with botanical material but 
four species. Four, incidentally mentioned in the text, he justly 
states can in all probability never be recognized, but the others 
he considers to represent characteristic species, which will 
eventually be recognized when the Moluccas shall have been 
more thoroughly explored botanically. 

Numerous questions of nomenclature cannot be determined 
from data at present available, and accordingly the following 
treatment of the palms described by Rumphius is distinctly un- 
satisfactory in many respects. Many of the more characteristic 
forms figured by him are readily recognizable, and their status 
is certain. Others must await the collection of additional ma- 
terial. Under the circumstances it has been considered best 
to give a critical enumeration of all the species, according to 
their definitely determined or their problematical positions. In 
certain cases new names are indicated, but few new combina- 

* Ann. Bot. Gard. Calcutta 11 (1908) 1-518, t. 1-238; 12 1 (1911) 1-237, 
t. 1-109. 



110 



RUMPHIUS'S HERBARIUM AMBOINENSE 



tions have been proposed in view of the uncertain status of some 
of the species. 

I am under obligations to Doctor 0. Beccari, Florence, Italy, 
for identifications of the palms actually collected by Doctor 
Robinson. 

CORYPHA Linnaeus 

CORYPHA UTAN Lam. Encycl. 2 (1786) 131 (type!). 

Borassus sylvestris Giseke Prael. Ord. Nat. PI. (1792) 86 (type!). 
Taliera sylvestris Blume ex Schultes Syst. 7 (1830) 1307 (type!). 
Corypha sylvestris Mart. Hist. Nat. Palm. 3 (1838) 233; Blume 

Rumphia 2 (1836) 59 (type!). 
Lontarus silvestris Rumph. Herb. Amb. 1: 53, t. 11. 

Lontarus silvestris Rumph. is the whole basis of Corypha utan 
Lam., which name should be maintained if the species proves to 
be a valid one. Borassus sylvestris Giseke, Corypha sylvestris 
Mart., and Taliera sylvestris Blume are exact synonyms, being, 
like Corypha utan Lam., based on Rumphius. 

CORYPHA GEBANGA Blume Rumphia 2 (1836) 59, t. 97, 89, 105; Mart. 
Hist. Nat. Palm. 3 (1838) 233. 
Gembanga rotundifolia Blume ex Nees in Flora 8 (1825) 580, 676. 
Taliera gembanga Blume ex Schultes Syst. 7 (1830) 1307. 
Lontarus silvestris s. cabang Rumph. Herb. Amb. 1:55. 

In this reduction I have followed Blume and Martius. In view 
of the uncertain status of Blume's species, no attempt is made 
to adjust the synonymy. It is manifestly very closely allied to, 
if not identical with, Corypha elata Roxb., and it may also prove 
to be the same as Corypha utan Lam. 

CORYPHA ELATA Roxb. Hort. Beng. (1814) 25, nomen nudum, Fl. Ind. 
ed. 2, 2 (1832) 176. 
Lontarus silvestris s. yhur e Philippin. Rumph. Herb. Amb. 1:54. 

Among all the palms that occur in the Philippines that yield 
a sago-like edible fecula, Rumphius's discussion of this applies 
only to Corypha in "trunco tarn crasso, ut vir eum brachiis 
complecti nequeat." Corypha elata Roxb. may prove to be iden- 
tical with one or both of the preceding species. Hasskarl, Neue 
Schliissel (1866) 12, quotes it as a possible synonym of Borassus 
sylvestris Gieseke, after Henschel, Vita Rumph. (1833) 140. 

LICU ALA Thunberg 

LICUALA RUM PHI I Blume Rumphia 2 (1836) 41, t. 89, /. 2. 

Corypha licuala Lam. Encycl. 2 (1783) 131 (type!). 
Licuala arbor Rumph. Herb. Amb. 1:44, t. 9. 

The status of this species, as distinct from Licuala spinosa 



PALMAE 



111 



Wurmb, in Verh. Bat. Genoets. 2 (1780) 469, is doubtful, and it 
may properly be a synonym of Wurmb's species. Loureiro, Fl. 
Cochinch. (1790) 213, placed it under Corypha pilearia Lour., 
which is Licuala pilearia Blume, and perhaps the same as Li- 
cuala spinosa Wurmb. It is the whole basis of Corypha licuala 
Lam. and in part the basis of Licuala rumphii Blume. Murray, 
Giseke, Roxburgh, Willdenow, Poiret, Schultes, and other au- 
thors refer it to Licuala spinosa Thunb., in Vet. Akad. Nya 
Handl. (1782) 278, which is antedated by two years by Licuala 
spinosa Wurmb. The essential distinctive characters of Licuala 
rumphii Blume are very imperfectly known, and a critical revi- 
sion of the genus may show that it is a synonym of Licuala 
spinosa Wurmb. The figure is poor, but it manifestly represents 
a Licuala very similar in appearance to the widely distributed 
Licuala spinosa Wurmb. 

LIVISTONA R. Brown 

LIVISTONA ROTUNDI FOLIA (Lam.) Mart. Hist. Nat. Palm. 3 (1838) 
241. 

Corypha rotundifolia Lam. Encycl. 2 (1786) 131 (type!). 
Saribus rotundifolius Blume Rumphia 2 (1836) 49. 
Licuala rotundifolia Blume ex Roem. & Schultes Syst. 7 2 (1830) 1305. 
Saribus Rumph. Herb. Amb. 1: 42, t. 8. 

This is one of the few Rumphian species considered by Lin- 
naeus in the first edition of his Species Plantarum (1753), 
where he erroneously reduced it to Corypha umbraculif era Linn., 
page 1187, and later cited it under the same name in Stickman 
Herb. Amb. (1754) 6, Amoen. Acad. 4 (1759) 118, and Sp. PL 
ed. 2, (1763) 1657, in whch he was followed by Giseke, Prael. 
Ord. Nat. PL (1792) 49. Louriero, FL Cochinch. (1790) 212, 
placed it under Corypha saribus Lour., taking his specific name 
from Rumphius. The species he actually described, however, 
is not the Amboina form, but is Livistona cochinchinensis Blume. 
Saribus Rumph. is the whole basis of Corypha rotundifolia Lam., 
which in turn typifies Livistona rotundifolia Mart. 

Arbor tsjang Rumph., Herb. Amb. 1: 63, which is very im- 
perfectly described, is perhaps a species of Livistona, as sug- 
gested by Hasskarl, Neue Schlussel (1866) 12; it was from 
French Indo-China. 

LIVISTONA ? BISSULA Mart. Hist. Nat. Palm. 3 (1838) 242 (type!). 
Licuala ? bissula Miq. Fl. Ind. Bat. 3 (1855) 57 (type!). 
Bissula Rumph. Herb. Amb. 1:85. 

The status of this species is entirely doubtful. It may be 



112 



RUMPHIUS'S HERBARIUM AMBOINENSE 



neither a Livistona nor a Licuala; the description of the fruits 
is suggestive of Pholido carpus. Rumphius's specimens were 
from Celebes, where the palm is known as bissula and metsje. 

PHOLIDOCARPUS Blume 
PHOLIDOCARPUS IHUR (Giseke) Blume Rumphia 3 (1837) 90 (type!). 
Borassus ? ihur Giseke Prael. Ord. Nat. PI. (1792) 87 (type!). 
Pholidocarpus rumphii Meisn. ex Hassk. in Abhandl. Naturf. Gesel- 

lsch. Halle 9 (1866) 154 (type!). 
Lontarus silvestris altera s. ihur Rumph. Herb. Amb. 1 : 56, t. 12. 

The Rumphian illustration and description typify Borassus 
ihur Giseke, Pholidocarpus ihur Blume, and Pholidocarpus 
rumphii Meisn. ; the last does not appear in Index Kewensis. 

BORASSUS Linnaeus 

BORASSUS FLABELLI FER Linn. Sp. PL (1753) 1187. 
Borassus flab elli for mis Murr. Syst. (1774) 827. 
Lontarus domestica Gaertn. Fruct. 1 (1788) 21, t. 8. 
Lontarus domestica Rumph. Herb. Amb. 1 : 45, t. 10. 

This was one of the few Rumphian species considered by 
Linnaeus in the first edition of his Species Plantarum (1753), 
where, page 1187, he correctly reduced it to Borassus flabellifer 
Linn., the reduction being repeated in his later writings, Stick- 
man Herb. Amb. (1754) 6, Amoen. Acad. 4 (1759) 118, Sp. 
PL ed. 2 (1763) 1657, and generally accepted by all authors. 
Gaertner, Fruct. 1 (1788) 21, merely substituted the generic- 
name Lontarus for the Linnean name Borassus, considering the 
species under the new name Lontarus domestica Gaertn. 

LODOICEA Commerson 

LODOICEA MALDIVICA (Gmel.) Pers. Syn. 2 (1807) 630. 

Cocos maldivica Gmel. Syst. (1791) 569 (maldioica), ed. 2 (1796) 
569. 

Borassus sonnerati Giseke Prael. Ord. Nat. PI. (1792) 86. 
Lodoicea callipyge Comm. ex St. Hil. Expos. 1 (1805) 96. 
Lodoicea seychellarum Labill. in Ann. Mus. Paris 9 (1807) 140, t. 13. 
Cocus maldivicus Rumph. Herb. Amb. 6: 210, t. 81. 

Rumphius figures only the fruit of this striking palm, and 
Gmelin's specific name, here adopted, is the oldest valid one for 
the species. 

METROXYLON Rottboell 

METROXYLON SAGU Rottb. in Nye Samml. Dansk. Vidensk. Selsk. Skrift. 
2 (1783) 527. 

Sagus genuina Giseke Prael. Ord. Nat. PI. (1792) 94 (type!). 
Metroxylon sagus Koenig in Spreng. Syst. 2 (1825) 138 (type!). 



PALMAE 



113 



Sagus rumphii Willd. Sp. PI. 4 (1805) 404 (type!). 

Metroxylon rumphii Mart. Hist. Nat. Palm. 3 (1838) 213. 

Sagus inermis Roxb. Hort. Beng. (1814) 68 (type!). 

Sagus genuina Rumph. Herb. Amb. 1: 72, t. 17, 18. 
Amboina, Paso, Robinson PI. Rumph. Amb. 26, October 28, 1913, locally 
known as sagu tuni. The inflorescence and fruit from a felled trunk, the 
leaves from a neighboring palm. Doctor Beccari, who has examined the 
material, thinks that perhaps the fruits belong with Metroxylon micra- 
canthum Mart. 

The sago palm presents considerable variation, especially in 
the presence or absence of spines, and when spines are present, 
in their length. It is probable that altogether too many 
"species" have been proposed and that most of the plants 
described by Rumphius, on which later authors have based 
various binomials, are really but forms or varieties of a 
single species. No material is available, however, for pur- 
poses of comparison, and accordingly the various species that 
have been based on Rumphius's descriptions are listed below 
with their proper synonyms. The Rumphian illustrations and 
descriptions, cited above, have been quoted under all of the 
synonyms above listed, and most of them have been based solely 
on Rumphius. I have here adopted what is manifestly the 
oldest specific name for the species. 

METROXYLON M ICRACANTH U M Mart. Hist. Nat. Palm. 3 (1838) 215 
(type!). 

Sagus micracanthus Blume Rumphia 2 (1836) 153 (type!). 
Sagus genuina II Rumph. Herb. Amb. 1: 75. 

This is probably merely a form of Metroxylon sagu Rottb., 
and it is possible that the fruit and spadices of the specimen 
cited under the above species belong here. 

METROXYLON LONGISPINUM Mart. Hist. Nat. Palm. 3 (1838) 215 
(type!). 

Sagus longispina Blume Rumphia 2 (1836) 154 (type!). 

Sagus genuina Giseke var. longispina Giseke Prael. Ord. Nat. PI. 

(1792) 94 (type!). 
Sagus longispina Rumph. Herb. Amb. 1: 75. 

This is apparently merely a form or variety of the common 
sago palm with long spines. However, no material is available 
to warrant its definite reduction. The native names cited by 
Rumphius are lapia macanaru, lapia macanalo, and lapia maca- 
nalun. It was reduced by Poiret, in Lam. Encycl. 6 (1804) 
394, to Sagus farinifera (Gaertn.) Lam., which is certainly not 
the correct disposition of it. 

144971 8 



114 



RUMPHIUS'S HERBARIUM AMBOINENSE 



METROXYLON INERME Mart. Hist. Nat. Palm. 3 (1838) 215. 
Sagus laevis Blume Rumphia 2 (1836) 147. 
Sagus laevis Rumph. Herb. Amb. 1: 76. 

Like the preceding this is probably only a form or variety 
of the common sago palm. It has been referred by various 
authors to Metroxylon laeve Mart., to Sagus spinosus Roxb., to 
Metroxylon hermaphroditum Hassk., and to Metroxylon sagus 
Rottb., besides the synonyms cited above. Native names cited 
by Rumphius are lapia molat, sagu parampuan, and bulun. 

METROXYLON SYLVESTRE Mart. Hist. Nat. Palm. 3 (1838) 215 
(type!). 

Sagus genuina Giseke var. silvestris Giseke Prael. Ord. Nat. PI. (1792) 

94 (type!). 
Sagus silvestris Rumph. Herb. Amb. 1: 75. 
Amboina, Paso, Robinson PL Rumph. Amb. 27, October 29, 1913, locally 
known as sagu ihur. 

The specimen has fairly long spines on its petioles, undoubtedly 
represents Sagus silvestris Rumph., and hence Metroxylon 
sylvestre Mart., but in all essential characters it appears to be 
identical with Metroxylon sagu Rottb. 

PIGAFETTIA Beccari 

PIGAFETTIA FILIFERA (Giseke) comb. nov. 

Sagus filifera Giseke Prael. Ord. Nat. PI. (1792) 94 (type!). 
Sagus filaris Blume Rumphia 2 (1836) 154 (type!). 
Metroxylon filare Mart. Hist. Nat. Palm. 3 (1838) 215 (type!). 
Pigafettia filaris Becc. Malesia 1 (1877) 90, in obs. 
Sagus filaris Rumph. Herb. Amb. 1: 84, t. 19. 

The species is still known only from the Rumphian description 
and figure, which are the type of Pigafettia filifera and of all the 
synonyms cited above. 

PIGAFETTIA ELATA (Reinw.) H. Wendl. in Kerch. Palm. (1878) 253. 
Sagus elata Reinw. ex Blume Rumphia 2 (1836) 156, t. 128, f. 1. 
Metroxylon elatum Mart. Hist. Nat. Palm. 3 (1838) 216. 
Wanga Rumph. Herb. Amb. 1 : 85. 

Wanga, a Celebesian palm, is very briefly described by 
Rumphius. Pigafettia elata H. Wendl. seems to be the correct 
disposition of it. 

ZALACCA Reinwardt 

ZALACCA EDULIS Reinw. ex Roem. & Schultes Syst. 7 2 (1830) 1334. 
Salacca edulis Reinw. Syll. Ratisb. 2 (1828) 3. 
Salakka edulis Reinw. ex Blume Cat. Gew. Buitenz. (1823) 112. 
Calamus zalacca Gaertn. Fruct. 2 (1791) 267, t. 139. 



PALMAE 



115 



Zalacca blumeana Mart. Hist. Nat. Palm. 3 (1836-50) 201, t. 123, 
t. 159, f. 3. 

Zalacca Rumph. Herb. Amb. 5: 113, t. 57. f. 2. 
Amboina, Koesoekoesoe sereh and Mahija, Robinson PI. Rumph. Amb. 
25, August 23, 1913, locally known as salak. 

Linnaeus, Sp. PL ed. 2 (1762) 463, originally reduced this 
through error to Calamus rotang Linn. Gaertner, Fruct. 2 
(1791) 267, cites it under Calamus zalacca Gaertn.= Zalacca 
edulis Reinw. Most authors have cited it under Zalacca (Sa- 
lacca) edulis Reinw., but Martius placed it under Zalacca blu- 
meana Mart., which is merely a synonym of Reinwardt's species. 

C A L A M US Linnaeus 

CALAMUS ALBUS Pers. Syn. 1 (1805) 383 (pp. excl. syn. Lour.) (type!); 
Beccari in Ann. Bot. Gard. Calcutta 11 (1908) 444, t. 199. 
Palmijuncus albus Rumph. Herb. Amb. 5: 102, t. 53. 
Amboina, Hitoe messen, Robinson PI. Rumph. Amb. 2J+, October 14, 1913, 
in forests, altitude 200 meters; locally known as rotang tuni. 

The specimen cited above has been identified with Calamus 
albus Pers. by Doctor Beccari. Palmijuncus albus Rumph. has 
been reduced by various authors to Calamus rotang Linn, and 
to Calamus rudentum Lour., most of the references in literature 
being to the latter species. Loureiro made the reduction to 
Calamus rudentum Lour, in the original description of that 
species, but the type was a Cochin-China specimen, and the 
species is certainly not the same as the Amboina one. Strictly 
following rules of nomenclature, the name Calamus albus Pers. 
belongs with Calamus rudentum Lour., as a synonym, but I have 
here followed Beccari in applying it to the Amboina plant. 

CALAMUS GRAMINOSUS Blume Rumphia 3 (1837) 31 (type!). 
Palmijuncus albus graminosus Rumph. Herb. Amb. 5: 104. 

A species known only from Rumphius's description; see Bec- 
cari in Ann. Bot. Gard. Calcutta 11 (1908) 499. It has been 
reduced by various authors to Calamus rotang Linn., C. rudentun 
Lour., and C. adspersus Blume, but is manifestly none of these. 
It may be a form of Calamus albus Pers. 

CALAMUS RUMPHII Blume Rumphia 3 (1837) 38 (type!). 

Daemonorops rumphii Mart. Hist. Nat. Palm. 3 (1850) 331. 
Palmijuncus verus angustifolius Rumph. Herb. Amb. 5: 105, t. 51f, f. 2. 

A species known only from Rumphius's figure and description ; 
see Beccari, in Ann. Bot. Gard. Calcutta 11 (1908) 400, who 
considers it to be allied to Calamus palustris Griff. It has been 
erroneously reduced by various authors to Calamus rotang Linn., 



116 



RUMPHIUS'S HERBARIUM AMBOINENSE 



to C. platyacanthos Mart., to Daemonorops elongatus Blume, and 
to D. longipes Griff. 

CALAMUS PISICARPUS Blume Rumphia 3 (1837) 31 (type!). 

Palmijuncus verus latifolius Rumph. Herb. Amb. 5: 106, t. 5U, /• 1- 

This species is known oniy from Rumphius's figure and descrip- 
tion, but is considered by Beccari, Ann. Bot. Gard. Calcutta 11 
(1908) 460, to represent a very distinct species allied to Cala- 
mus aruensis Becc. and C. holrungii Becc. By various authors 
it has been erroneously reduced to Calamus rotang Linn., to 
C. verus Lour., to C. oblongus Reinw., and to C. strictus Blume. 
Manifestly Calamus verus Lour., described from Cochin-China 
specimens, is a species entirely different from the Amboina plant, 
although Loureiro cites Palmijuncus verus Rumph. as a synonym 
of his species and takes his specific name from Rumphius. 

CALAMUS VIMINALIS Willd. Sp. PL 2 (1799) 203 (type!). 
Calamus litoralis Blume Rumphia 3 (1837) 43. 
Palmijuncus viminalis Rumph. Herb. Amb. 5: 108, t. 55, /. 2. 

I agree with Beccari, Ann. Bot. Gard. Calcutta 11 (1908) 
203, 497, that the Rumphian illustration pertains to the Javan 
species, not to the form described from Buru (Calamus buroensis 
Mart.). Palmijuncus viminalis Rumph. typifies Calamus vimi- 
nalis Willd., and it has been erroneously reduced by various 
authors to Calamus rotang Linn., to C. fasciculatus Roxb., and 
to C. buroensis Mart. 

CALAMUS BUROENSIS Mart. Hist. Nat. Palm. 3 (1850) 336 (type!). 

Calamus viminalis e Burone Rumph. Herb. Amb. 5: 109 (non t. 55, 
/■ *)- 

This is a species of doubtful status, known only from Rumph- 
ius's description; see Beccari in Ann. Bot. Gard. Calcutta 11 
(1908) 497. 

CALAMUS EQU ESTRIS Willd. Sp. PI. 2 (1799) 204 (type!). 
Palmijuncus equestris Rumph. Herb. Amb. 5: 110, t. 56. 

A species known only from Rumphius's figure and description ; 
see Beccari in Ann. Bot. Gard. Calcutta 11 (1908) 358. It has 
been erroneously reduced by various authors to Calamus rotang 
Linn, and to C. dioicus Lour. Willdenow also refers here 
t. 57, f. 1, which Beccari places under Calamus caiva Blume. 

CALAMUS CAWA Blume Rumphia 3 (1837) 31, 62 (type!). 

Palmijuncus equestris s. rottang cawa Rumph. Herb. Amb. 5: 112, 
t. 57, f. 1. 

A species known only from Rumphius's description; see Bee- 



PALMAE 



117 



cari in Ann. Bot. Gard. Calcutta 11 (1908) 357. It has been 
confused by various authors with Calamus equestris Willd., with 
C. javensis Blume, and with C. maritimus Blume. 

CALAMUS AC ID US Becc. in Ann. Bot. Gard. Calcutta 11 (1908) 496 
(type!). 

Palmijuncus acidus Rumph. Herb. Amb. 6: 119, t. 58, f. 2, E. 

A very imperfectly known species, known only from Rumph- 
ius's description and figure. By various authors it has been 
erroneously reduced to Calamus barbatus Blume; to Calamus 
oblongus "Linn.," which does not exist [Hasskarl, Neue Schliissel 
(1866) 102] ; and to Daemonorops barbatus Mart. 

CALAMUS sp. 

Palmijuncus aracanicus Rumph. Herb. Amb. 5: 107. 

A form of entirely uncertain status ; it was thought by Schultes 
to be referable to Calamus oblongus Reinw., and by Kunth to be 
Calamus latifolius Roxb. 

CALAMUS sp.? 

Palmijuncus viminalis s. ua huay Rumph. Herb. Amb. 5: 109. 

Entirely indeterminable from any data given by Rumphius; 
it may not be a Calamus at all. 

DAEMONOROPS Blume 

DAEMONOROPS CALAPPARIUS Blume Rumphia 3 (1837) 7 (type!). 
Calamus calapparius Mart. Hist. Nat. Palm. 3 (1850) 331. 
Calamus amboinensis Miq. in Verh. Kon. Akad. Wetensch. 11 (1868) 
20. 

Palmijuncus calapparius Rumph. Herb. Amb. 5: 98, t. 51. 
Amboina, Lateri, Robinson PL Rumph. Amb. 23, September 5, 1913, in 
forests, altitude about 250 meters, locally known as rotang bulu rusa. 

The identification of this specimen with both Palmijuncus 
calapparius Rumph. and Daemonorops calapparius Blume has 
been made by Doctor Beccari, who has given a detailed descrip- 
tion of the species from Amboina specimens.* Linnaeus 
originally reduced this to Calamus rotang Linn., Sp. PL ed. 2 
(1762) 463, in which he was followed by numerous authors. 
Loureiro, Fl. Cochinch. (1790) 209, reduced it to Calamus 
petraeus Lour., but the Cochin-China species actually described 
by Loureiro is entirely different from the Amboina plant, and 
Beccari thinks it probably a species of Korthalsia or Plectocomia. 



*Ann. Jard. Bot. Calcutta 12 1 (1911) 164. 



118 



RUMPHIUS'S HERBARIUM AMBOINENSE 



DAEMONOROPS NIGER (Willd.) Blume Rumphia 3 (1827) 5 (type!). 
Calamus niger Willd. Sp. PI. 2 (1799) 203 (type!). 
Palmijuncus niger Rumph. Herb. Amb. 5: 101, t. 52. 

A species known only from the Rumphian figure and descrip- 
tion; see Beccari in Ann. Bot. Gard. Calcutta 12 1 (1911) 104. 
Palmijuncus niger Rumph. has been reduced by various au- 
thors to Calamus rotang Linn., to Calamus rudentum Lour., and 
to Daemonorops melanochaetes Blume, but these reductions are 
manifestly erroneous. 

DAEMONOROPS PAL EM BAN ICUS Blume Rumphia 3 (1837) 20. 
Palmijuncus palimbanicus Rumph. Herb. Amb. 5: 107. 

The reduction follows Blume, which is probably the correct 
disposition of Rumphius's species. 

DAEMONOROPS DRACO (Willd.) Blume Rumphia 3 (1837) 8. 
Calamus draco Willd. Sp. PI. 2 (1799) 203 (type!). 
Palmijuncus draco Rumph. Herb. Amb. 5: 114, t. 58, f. A, B. 

Regarding this species, Beccari, Ann. Bot. Gard. Calcutta 12 1 
(1911) 106, states: "Only the Palmijuncus Draco growing at 
Palembang, described and figured by Rumphius (1. c.) can be 
considered as the true D. Draco," and again: "The canes 
described by Rumphius * * * figured in plate 58 f. D 
* * * and which is there attributed to Palmijuncus Draco, 
are almost certainly those of Calamus Scipionum." 

DAEMONOROPS RUBER Blume Rumphia 3 (1837) 6. 
Daemonorops accedens Blume 1. c. 13. 

Palmijuncus draco e Bantam Rumph. Herb. Amb. 5: 116. 

The Javan material included by Rumphius in his description 
of Palmijuncus draco was referred by Blume to Daemonorops 
accedens Blume, which is a synonym of D. ruber Blume; see 
Beccari in Ann. Bot. Gard. Calcutta 12 1 (1911) 114. 

CARYOTA Linnaeus 

CARYOTA RUMPH I AN A Mart. Hist. Nat. Palm. 3 (1839) 195 (type!). 
Saguaster major Rumph. Herb. Amb. 1: 64, t. H. 

The illustration is an excellent habit sketch of this sufficiently 
well-known palm, which extends from the Philippines to the 
Moluccas and Java. Saguaster major Rumph. is the whole basis 
of Caryota rumphiana Mart., from which Caryota maxima Blume 
does not appear to be distinct. Linnaeus originally reduced the 
Rumphian species to the Indian Caryota urens Linn., in Stick- 
man Herb. Amb. (1754) 6, Amoen. Acad. 4 (1759) 118, and 



PALMAE 



119 



it is very generally cited as a synonym of the Linnean species 
in the early botanical literature. The species is, however, en- 
tirely distinct from Caryota wrens Linn. 

ARENGA * LaBillardiere 

ARENGA PI N N ATA (Wurmb) comb. nov. 

Saguerus pinnatus Wurmb in Verh. Bat. Genoots. 1 (1779) 351. 

Borassus gomutus Lour. Fl. Cochinch. (1790) 618. 

Arenga saccharifera Labill. in Mem. Inst. Paris 4 (1801) 209. 

Gomutus rumphii Corr. in Ann. Mus. Paris 9 (1807) 288. 

Saguerus rumphii Roxb. Hort. Beng. (1814) 68 (type!). 

Saguerus saccharifer Blume Rumphia 2 (1836) 128. 

Gomutus saccharifer Spreng. Syst. 2 (1825) 624. 

Arenga gamuto Merr. in Philip. Journ. Sci. 9 (1914) Bot. 63. 

Palma indica vinaria II Rumph. Herb. Amb. 1:57, t. 13. 

The common and well-known sugar palm is not represented 
in our Amboina collections. The figure and the description 
given by Rumphius unmistakably represent the form commonly 
known as Arenga saccharifera Labill. The illustration has been 
cited under Borassus gomutus Lour., Saguerus rumphii Roxb., 
Saguerus saccharifer Blume, and Arenga saccharifera Labill., 
all synonyms of Arenga pinnata (Wurmb) Merr. ; while Giseke, 
Prael. Ord. Nat. PL (1792) 90, erroneously reduced it to 
Chamaerops humilis Linn. Wurmb's specific name is the oldest 
valid one for the species. Arenga gamuto Merr. was adopted 
on the assumption that ''Saguerus gamuto Houtt." was published 
as indicated in Index Kewensis and in the synonymy of Arenga 
saccharifera Labill. as given by some authors; no such name 
appears in Houttuyn's work. 

CALYPTROCALYX Blume 

CALYPTROCALYX SPICATUS (Lam.) Blume Rumphia 2 (1836) 103, 
t. 102, f. 2, 118, 161. 
Areca spicata Lam. Encycl. 1 (1783) 241 (type!). 
Pinanga silvestris globosa Rumph. Herb. Amb. 1: 38, t. 5, f. 1, A. 

This species is not represented in our Amboina collections, 
but it has been minutely described and figured by Blume from 
Amboina specimens. The Rumphian description and crude fig- 
ure are the whole basis of Areca spicata Lam. Gaertner, Fruct. 
1 (1788) 24, referred it to Euterpe glohosa Gaertn., in which he 
was followed by Giseke, Prael. Ord. Nat. -PL (1792) 92, but the 
fruit figured by Gaertner is not that of Calyptrocalyx spicatus 
Blume. Teysmann, quoted by Hasskarl, Neue Schliissel (1866) 



* Retained name, Vienna Code; Saguerus Adans. (1763) is older. 



120 RUMPHIUS'S HERBARIUM AMBOINENSE 



10, thought that it was a species of Iguanura, while by other 
authors it has been placed as an undetermined species of Areca 
and of Pinanga. 

DRYMOPHLOEUS Zippel 

DRYMOPHLOEUS OLIVAEFORMIS (Giseke) Mart. Hist. Nat. Palm. 3 
(1849) 314. 

Areca olivaeformis Giseke Prael. Ord. Nat. PI. (1792) 79 (type!). 
Harina rumphii Mart. Hist. Nat. Palm. 189, ex Kunth. Enum. 3 
(1841) 194. 

Seaforthia olivaeformis Mart. Hist. Nat. Palm. 3 (1849) 314. 

Ptychosperma rumphii Blume Rumphia 2 (1836) 119. 

Saguaster minor Rumph. Herb. Amb. 1: 67, t. 15. 
Amboina, Lateri and Gelala, Robinson PI. Rumph. Amb. 21, 22, August 
and September, 1913, in ravines and in forests, altitude 20 to 200 meters, 
locally known as pinang utan. 

Saguaster minor Rumph. is the basis of Areca olivaeformis 
Giseke, this supplying the oldest valid specific name. Hamilton, 
Mem. Wern. Soc. 5 (1826) 316, discusses it under Harina caryo- 
toides Ham., which is described from Indian specimens, but 
he does not reduce Saguaster minor to this species. 

DRYMOPHLOEUS APPENDICULATUS Becc. Malesia 1 (1877) 46. 
Areca gracilis Giseke Prael. Ord. Nat. PI. (1792) 80 (type!). 
Ptychosperma appendiculata Blume Rumphia 2 (1836) 122, t. 8U, 119. 
Seaforthia jaculatoria Mart. Hist. Nat. Palm. 3 (1849) 314 (type!). 
Drymophloeus jaculatorius Mart. Hist. Nat. Palm. 3 (1849) 314 
(type!). 

Sargile Rumph. Herb. Amb. 1: 68. 

Areca gracilis Giseke was based wholly on Rumphius's descrip- 
tion of Sargile. Beccari, Malesia 1 (1877) 46, has excluded it 
from the synonymy of Drymophloeus appendiculatus (Blume) 
Becc, with all other synonyms based on Rumphius, because of the 
fact that Rumphius included in the description more than one 
species, and because he based his description largely on data 
supplied to him by the natives. Beccari, op. cit. 98, refers 
"Saguaster minor ex Gilolo et Nova-Guinea/' i. e., Sargile 
Rumph. p. p., to Drymophloeus jaculatorius Mart., a species of 
doubtful status. 

DRYMOPHLOEUS sp. 

Areca vaginata Giseke Prael. Ord. Nat. PI. (1792) 80 (type!). 
Sargile minor II Rumph. Herb. Amb. 1: 68. 

The form described in the last paragraph of chapter fifteen was 
made the type of Areca vaginata by Giseke. It may prove to 
be a synonym of the preceding species, but its status is now 
entirely doubtful. 



PALMAE 



121 



DRYMOPHLOEUS ? sp. 

Areca humilis Willd. Sp. PI. 4 (1805) 595 (type!). 
Seaforthia saxatilis Blume ex Mart. Hist. Nat. Palm. 3 (1838) 186 
(type!). 

Ptychosperma saxatilis Blume Rumphia 2 (1836) 121 (type!). 
Pinanga silvestris saxatilis Rumph. Herb. Amb. 1: 42, t. 7. 

This species is not represented in our Amboina collections; 
and, although it has been placed in several different genera, its 
proper place is still uncertain. Loureiro, FL Cochinch. (1790) 
619, discusses it under Borassus caudata Lour., but does not 
definitely refer it to this species. The Rumphian figure and 
description are the whole basis of Areca humilis Willd., which 
thus supplies the oldest specific name for the species when once 
its proper genus is determined. It is also the whole basis of 
Ptychosperma saxatilis Blume and of Seaforthia saxatilis Blume. 
"Areca saxatilis Burm.," FL Ind. (1768) 42, quoted by Miquel 
and listed in Index Kewensis, does not appear in Burman's work. 
Beccari, Malesia 1 (1877) 47, excludes this from Drymophloeus, 
but make no suggestion as to its proper disposition. 

ACTI NORH YTIS H. Wendland and Drude 

ACTINORH YTIS CALAPPARIA (Blume) H. Wendl. & Drude in Linnaea 
39 (1875) 184. 
Areca calapparia Blume Rumphia 2 (1836) 70, t. 100, f. 2. 
Seaforthia calapparia Mart. Hist. Nat. Palm. 3 (1849) 313. 
Ptychosperma ? calapparia Miq. PI. Ind. Bat. 3 (1855) 20. 
Pinanga calapparia Rumph. Herb. Amb. 1:28. 

This species is not represented in our Amboina collections. 
The reduction made by Blume and followed by Miquel and by 
Martius is apparently the correct disposition of Pinanga calap- 
paria Rumph. 

M ISCHOPH LOEUS Scheffer 

MISCHOPHLOEUS VESTIARIA (Giseke) comb. nov. 

Areca vestiaria Giseke Prael. Ord. Nat. PI. (1792) 78 (type!). 
Seaforthia vestiaria Mart. Hist. Nat. Palm. 3 (1849) 313 (type!). 
Ptychosperma vestiaria Miq. PI. Ind. Bat. 3 (1855) 31 (type!). 
Ptychosperma paniculata Miq. in Verh. Kon. Akad. Wetensch. 1 1 
(1868) 3. 

Areca paniculata Scheff. in Naturk. Tijdschr. Ned. Ind. 32 (1873) 
168. 

Mischophloeus paniculata Scheff. in Ann. Jard. Bot. Buitenz. 1 (1876) 
152. 

Pinanga silvestris e Buro Rumph. Herb. Amb. 1:41. 

I have here adopted what is apparently the oldest specific 
name for this species, as Areca vestiaria Giseke was based wholly 



122 



RUMPHIUS'S HERBARIUM AMBOINENSE 



on Rumphius's description, as were Seaforthia vestiaria Mart, 
and Ptychosperma vestiaria Miq. There seems to be no doubt 
that Mischophloeus paniculata (Miq.) Scheff. is the same species 
as the plant Rumphius described. 

PINANGA Blume 

PINANGA PUNICEA (Blume) comb. nov. 

Areca punicea Blume Rumphia 2 (1826) 72 (type!). 
Seaforthia rumphiana Mart. Hist. Nat. Palm. 3 (1849) 314 (type!). 
Drymophloeus rumphianus Mart. Hist. Nat. Palm. 3 (1849) 314 
(type!). 

Ptychosperma punicea Miq. Fl. Ind. Bat. 3 (1855) 31 (type!). 
Pinanga ternatensis Scheff. in Ann. Jard. Bot. Buitenz. 1 (1876) 149. 
Pinanga silvestris glandiformis II Rumph. Herb. Amb. 1:39. 

Areca punicea Blume is based wholly on Rumphius's descrip- 
tion of Pinanga silvestris glandiformis II, as are the brief de- 
scriptions of Drymophloeus rumphianus Mart. {Seaforthia 
rumphiana Mart.) and Ptychosperma punicea Miq. The species 
is undoubtedly a Pinanga and is probably the species described 
by Scheffer as Pinanga ternatensis. I have here adopted Blume's 
specific name, it being the oldest valid one. Sarasuac Camell, 
cited by Blume as a synonym of Areca punicea Blume, Rumphia 
2 (1836) 73, is Heterospathe elata Scheff., a species originally 
described from Amboina material, and one that I cannot connect 
with any form described by Rumphius. 

PINANGA GLOBULI FERA (Lam.) comb. nov. 

Areca globulifera Lam. Encycl. 1 (1783) 241 (type!). 
Areca ory zaef ormis Giseke Prael. Ord. Nat. PI. (1792) 76 (type!). 
Pinanga silvestris oryzaeformis Rumph. Herb. Amb. 1: 40, t. 5, f. 2, 
B, C, D. 

Nothing resembling the Rumphian species is represented in 
our Amboina collections, although the figure apparently repre- 
sents a species of Pinanga. Pinanga silvestris oryzaeformis 
Rumph. is the whole basis of Areca globulifera Lam., which has 
been reduced to Pinanga kuhlii Blume, and which, if correctly 
placed, supplies a much older specific name for that species. 
Loureiro, Fl. Cochinch. (1790) 568, reduced the Rumphian plant 
to Areca sylvestris Lour., but the Cochin-China plant that he 
actually described is certainly not the same as the one Rumphius 
described and figured. Gaertner, Fruct. 1 (1788) 20, referred 
it to Areca oryzaeformis Gaertn., and the fruit he figured is 
probably that of Pinanga kuhlii Blume. Seaforthia oryzaefor- 
mis Mart., Hist. Nat. Palm. 3 (1839) 185, is merely a transfer 



PALMAE 



123 



of Areca oryzaeformis Gaertn., to be typified by the plant 
Gaertner actually figured; rather than by the Rumphian 
synonym he cites and from which he took the specific name. 

ARECA Linnaeus 

ARECA CATECHU Linn. Sp. PI. (1753) 1189 (err. cathecu) . 

Pinanga (incl. II alba et III nigra) Rumph. Herb. Amb. 1: 26, t. U- 

The common betel-nut palm is not represented in our Amboina 
collections. The reduction was first made by Linnaeus, in Stick- 
man Herb. Amb. (1754) 6, Amoen. Acad. 4 (1759) 118, which 
has been accepted by all authors, as it is manifestly the correct 
disposition of the form figured. Under this species Rumphius 
described four forms: namely, I Pinanga calapparia, which is 
apparently Actinorhytis calapparia H. Wendl. & Drude; 77 Pi- 
nanga alba and 777 Pinanga nigra, which are manifestly Areca 
catechu Linn.; and IV (unnamed), which I cannot place from 
the imperfect description given by him. 

ARECA GLAN Dl FORM IS Lam. Encycl. 1 (1783) 241 (type!). 

Pinanga silvestris glandiformis II Rumph. Herb. Amb. 1: 38, t. 6. 

The Rumphian figure and description are the whole basis of 
Areca glandiformis Lam. It has been minutely described and 
figured by Blume, Rumphia 2 (1836) 73, t. 100, from Moluccan 
material. 

COCOS Linnaeus 

COCOS NUCIFERA Linn. Sp. PI. (1753) 1188. 

Palma indica nucifera major s. calappa Rumph. Herb. Amb. 1:1, 
10, 11, 12, tt. 1-3. 

Rumphius gives an extensive treatise on the coconut. The 
illustrations present a habit sketch, infructescence and infloresc 
ence, details of the structure of the fruit, including germination, 
and t. 3 an abnormal form of the palm. He characterizes 
thirteen forms, most of which are referred by Blume, Miquel, 
and Hasskarl to various named varieties of Cocos nucifera 
Linn. The reduction of the illustrations was made by Linnaeus, 
in Stickman Herb. Amb. (1754) 6, Amoen. Acad. 4 (1759) 118, 
which is manifestly the correct disposition of them, and which 
has been accepted by all authors. While many distinct forms 
and varieties of the coconut occur, a clear understanding of them 
and their relationships is possible only through critical and long- 
continued field work, so that little is to be gained in attempting 
to reduce the numerous Rumphian forms to named and very 
imperfectly described varieties. 



124 



RUMPHIUS'S HERBARIUM AMBOINENSE 



N I PA (NYPA) Wurmb 

NIPA (NYPA) FRUTICANS Wurmb Verh. Bat. Genoots. 1 (1779) 350. 
Nypa Rumph. Herb. Amb. 1: 69, t. 16. 

The illustration is an excellent one of the common nipa palm, 
and Nypa was reduced by Wurmb in the original publication of 
Nipa (Nypa) fruticans Wurmb. This is manifestly the correct 
disposition of it and has been accepted by all authors, although 
commonly appearing in literature as Nipa fruticans Thunb. 

PALMAE OF UNCERTAIN STATUS 

I am unable satisfactorily to determine the status of the few forms 
described by Rumphius immediately following Cocus maldivicus. They may 
or may not be the seeds or fruits of palms. These doubtful forms are as 
follows : 

Compar mangae Rumph. Herb. Amb. 6: 216, t. 82, f. 1. 
Cocus maldivicus minor Rumph. Herb. Amb. 6: 218, t. 82, f. 2, S. 
Cocus melindanus verus Rumph. Herb. Amb. 6: 219, t. 82, f. If. 
Calapput laut Rumph. Herb. Amb. 6: 219. 

ARACEAE 

POT H OS Linnaeus 

POTHOS RUM PHI I Schott Melet. 1 (1832) 21. 

Scindapsus rumphii Presl Epim. (1851) 241. 

Adpendix porcellanica Rumph. Herb. Amb. 5: 485, t. 182, f. 1. 
Amboina, Paso, Robinson PI. Rumph. Amb. 116, 117, October, 1913, on 
trees at low altitudes, locally known as tapinawa and tapinawa puti. 

The Rumphian figure is cited by Presl, under Scindapsus 
rumphii Presl, and presumably is cited by Schott in the original 
publication of the species (not seen by me) . The species is a 
very characteristic and strongly marked one, the specimens cited 
above agreeing well with the Rumphian figure and description 
and with Pothos rumphii Schott as currently interpreted. 

POTHOS LONGIFOLIUS Presl Epim. (1851) 242. 

Adpendix duplo folio Rumph. Herb. Amb. 5: 490, t. 18U, f. 1-3. 
Amboina, Hitoe lama, Robinson PL Rumph. Amb. 118, November, 1913, 
on trees, altitude about 50 meters. 

Linnaeus originally reduced Adpendix duplo folio Rumph. to 
Pothos scandens Linn., which is very closely allied to Pothos 
longifolius Presl, and extends from India to Indo-China, Java, 
and Borneo. Pothos longifolius Presl is known from the Philip- 
pines, Sumatra, Java, and the Moluccas. The original reduc- 



ARACEAE 



125 



tion was made in Stickman, Herb. Amb. (1754) 25, Amoen. 
Acad. 4 (1759) 133, Syst. ed. 10 (1759) 1252, Sp. PL ed. 2 
(1763) 1374, which was followed by most of the early authors. 
Loureiro, Fl. Cochinch. (1790) 212, placed it under Flagellaria 
repens Lour. =Pothos loureirii Hook. & Arn. Hasskarl, Neue 
Schliissel (1866) 150, has interpreted the three forms described 
and figured by Rumphius as follows: I m&joT=Scindapsus 
officinalis Schott, most certainly wrong; II minor —Pothos rox- 
burghii DeVriese with question, but most certainly wrong, as 
the species is known only from India; and III=P. roxburghii 
DeVriese, with doubt. I can see no valid reason for considering 
that more than a single species is represented, and that is Pothos 
longifolius Presl. 

POTHOS LATIFOLIUS Linn, in Stickman Herb. Amb. (1754) 25, Amoen. 
Acad. 4 (1754) 133, Syst. ed. 10 (1759) 1252 (type!). 
Pothos tener Schott Aroid. 1 (1853) 24 (type!). 
Scindapsus tener Presl Epim. (1851) 241. 
Pothos gracilis Roxb. Fl. Ind. ed. 2, 1 (1832) 433. 
Adpendix arborum Rumph. Herb. Amb. 5: 483, t. 181, f. 1, 2. 

This is not represented in our Amboina collections and is a 
species of doubtful status. Pothos latifolius Linn, was based 
wholly on Rumphius's description and figure, the original de- 
scription being as follows: "181. Adpendix arborum = Pothos 
latifolius, foliis ovatis, petiolo latioribus." Thus Linnaeus in- 
cluded both figures 1 and 2, that is, the forms indicated by 
Rumphius as parvifolia and media. To me they both appear to 
represent the same species, one in, or immediately following, 
anthesis; the other in fruit. Pothos tener Schott was based 
wholly on Adpendix arborum I Rumph. Herb. Amb. 5: 483, t. 
181, f. 1; and, if I am correct in my surmise that but a single 
species is represented by the two figures, it becomes a synonym 
of Pothos latifolius Linn. Engler considers that Pothos tener 
Schott is apparently the same as the Bornean Pothos barberianus 
Schott. Pothos gracilis Roxb. was' described from specimens 
grown in Calcutta, originating in Amboina, and is probably a 
synonym of Pothos latifolius Linn.; Roxburgh states that it 
had a great resemblance to Adpendix arborum as figured by 
Rumphius. It is not, however, mentioned in the latest mono- 
graph of the group, Engl. Pflanzenreich 21 (1905) 21-44, and 
is quite certainly not the Penang and Bornean form described 
under the name Pothos gracilis Roxb. by Engler in DC. Monog. 
Phan. 2 (1879) 91. 



126 



RUMPH IUS'S HERBARIUM AMBOINENSE 



SCINDAPSUS Schott 

SCINDAPSUS M ARANTI FOLI US Miq. Fl. Ind. Bat. 3 (1857) 187. 
Cuscuaria mar anti folia Schott Gen. Aroid. (1858) t. 80. 
Cuscuaria rumphii Schott in Ann. Mus. Bot. Lugd. Bat. 1 (1863) 
130. 

Scindapsus cannaeformis Engl, in Bull. Soc. Tosc. Ort. 4 (1879) 
271. 

Aglaonema cuscuaria Miq. Fi. Ind. Bat. 3 (1857) 217 (type!). 
Scindapsus cuscuaria Engl. & Krause in Engl. Pflanzenreich 37 

(1908) 68, non Pothos cuscuaria Aubl. 
Pothos cuscuaria Gmel. Syst. 1 (1796) 274 (type!), non Aubl. 1775. 
Adpendix cuscuaria latifolia Rumph. Herb. Amb. 5: 488, t. 183, /. 1. 
Amboina, Soja, Way tommo, and near the town of Amboina, Robinson 
PL Rumph. Amb. 114, 115, August, 1913, on trees, altitude 7 to 250 meters, 
locally known as kelady utan and daun mo. 

The specific name cuscuaria is invalid for this species. Pothos 
cuscuaria Aubl., Hist. PL Guiane Franc. (1775) 840, was based on 
specimens from South America with no reference to Rumphius, 
although the specific name may have been taken from Rumphius. 
Pothos cuscuaria Gmel., Syst. 1 (1796) 274, however, was based 
wholly on Rumphius, not on Aublet. Aglaonema cuscuaria Miq. 
was based on Pothos cuscuaria Gmel. (non Aubl.) and Cuscua- 
ria latifolia Rumph., the latter being, accordingly, its type. 
Poiret, in Lam. Encycl. 5 (1804) 605, follows Gmelin, as do also 
Roemer and Schultes, Kunth, and Dietrich. 

ACORUS Linnaeus 

ACORUS CALAMUS Linn. Sp. PL (1753) 324. 

Acorum palustre et terrestre Rumph. Herb. Amb. 5: 178, t. 72, f. 1. 

The common sweet flag, amply described and fairly well 
figured by Rumphius, is not represented in our Amboina col- 
lections. It is found widely scattered in the Malayan region, 
sometimes cultivated, occasionally wild at higher altitudes. It 
was first reduced by Linnaeus in Stickman, Herb. Amb. (1754) 
20, Amoen. Acad. 4 (1759) 130, which has very generally been 
followed by later authors. Loureiro, Fl. Cochinch. (1790) 208, 
placed it under Orontium cochinchinense Lour. = Acorns cochin- 
chinensis Schott- = A corns calamus Linn. Schultes, Syst. 7 1 
(1829) 174, retains Acorus terrestris Rumph. as a species dis- 
tinct from A. calamus Linn., following Spreng., Syst. 2 (1825) 
118, who in turn based Acorus terrestris Spreng. on Acorus 
calamus Lour. Engler, DC. Monog. Phan. 2 (1879) 217, cites 
the Rumphian figure and description under Acorus calamus 
Linn. var. terrestris (Spreng.) Engl. 



ARACEAE 



127 



EPIPREMNUM Schott 

EPIPREMNUM PINNATUM (Linn.) Engl. Pflanzenreich 37 (1908) 60. 

Pothos pinnata Linn. Sp. PI. ed. 2 (1763) 1374 (type!). 

Scindapsus pinnatus Schott Melet. 1 (1832) 21. 

Adpendix laciniata Rumph. Herb. Amb. 5: 489, t. 183, f. 2. 
Amboina, Ayer putri and Amahoesoe, Robinson PI. Rumph. Amb. 113, 
August, 1913, on trees at low altitudes, locally known as tapinawa. 

Pothos pinnata Linn., as originally published, is based wholly 
on the description and figure of Rumphius's Adpendix laciniata. 
In addition to the synonyms given above, it has been cited by 
some authors under Rhaphidophora lacera Hassk. and Scindapsus 
pertusus Schott, both synonyms of Epipremnum pinnatum Engl. 
The species extends from India through Malaya to Polynesia. 

AMORPHOPHALLUS * Blume 

AMORPHOPHALLUS CAMPANULATUS (Roxb.) Blume ex Decne. in 
Ann. Mus. Hist. Nat. Paris 3 (1834) 366. 
Arum campanulatum Roxb. Hort. Beng. (1814) 65, nomen nudum, PI. 

Coromandel 3 (1819) 69, t. 272, Fl. Ind. ed. 2, 3 (1832) 509. 
Arum rumphii Gaudich. Bot. Freyc. Voy. (1826) 427, i. 39. 
Amorphophallus sativus Blume Rumphia 1 (1835) 145 (type!). 
Conophallus ? sativus Schott Prodr. (1860) 35. 
Tacca sativa Rumph. Herb. Amb. 5: 324, t. 112. 

Tacca phallifera Rumph. quoad Taccae fungus Rumph. Herb. Amb. 
5: 326, t. 113, f. 2. 

Not represented in our Amboina collections. Tacca sativa 
Rumph. is perhaps a mixture of Tacca pinnatifida Forst. and 
Amorphophallus; Tacca phallifera Rumph. certainly is. The 
description of the vegetative characters, however, manifestly 
applies to Amorphophallus campanulatus Blume; the figure is 
very poor. Linnaeus erroneously reduced Tacca sativa Rumph. 
to Dracontium polyphyllum Linn., in Stickman Herb. Amb. (1754) 
22, Amoen. Acad. 4 (1759) 131. Loureiro, Fl. Cochinch. (1790) 
300, placed it under Tacca pinnatifida Forst. Roxburgh cor- 
rectly placed it under Arum campanulatum Roxb. = Amorpho- 
phallus campanulatus, together with the inflorescence described 
as Taccae fungus and figured, 1. 113, f. 2. Amorphophallus sativus 
Blume is based wholly on the Rumphian description and figure, 
and this species Engler, Pflanzenreich 48 (1911) 109, includes 
under species dubiae. 

Tacca phallifera Rumph. is made up of Tacca pinnatifida 
Forst. and the inflorescence of Amorphophallus campanulatus 

* Retained name, Brussels Congress; Candarum Reichb. (1832) is older. 



128 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Blume, the plate presenting the vegetative parts of Tacca, its 
infructescence, and detached fruits, /. 1, a, b; and /. 2 the in- 
florescence of Amorphophallus campanulatus Blume, described 
by Rumphius as Taccae fungus. The latter figure is cited by 
Roxburgh in the original description of Arum campanulatum, 
and is certainly correctly placed. 

HOMALOMENA Schott 

HOMALOMENA COR DATA (Houtt.) Schott Melet. 1 (1832) 20. 

Dracontium cordatum Houtt. Handleid. 11 (1774-83) 200, t. 71, f. 2. 
Dracunculus amboinicus niger Rumph. Herb. Amb. 5: 322, t. Ill, 
f. 2. 

Amboina, Batoe merah, Robinson PI. Rumph. Amb. Ill, August 24, 1913, 
on shaded banks, altitude about 15 meters. 

Dracunculus amboinicus, as figured by Rumphius, was erro- 
neously reduced by Linnaeus to Arum divaricatum Linn., in 
Stickman Herb. Amb. (1754) 22, Amoen. Acad. 4 (1759) 131. 
By Loureiro, Fl. Cochinch. (1790) 532, it was placed with 
doubt under Calla occulta Lour., which may be the same as the 
closely allied Homalomena aromatica Schott. By other authors 
it has been reduced to Homalomena rubescens Kunth, H. rubra 
Kunth, and other allied forms. The Amboina specimen cited 
above agrees with the figure given by Rumphius, agrees with 
the description of Dracunculus amboinicus niger, and agrees 
also with Homalomena cordata Schott as recently described by 
Engler, Pflanzenreich 55 (1912) 57. Rumphius does not indicate 
which of the four forms described under Dracunculus amboinicus 
he intended the figure to represent, but I agree with Engler in 
referring it to the one first described, namely, Dracunculus 
amboinicus niger. 

The form described by Rumphius as Dracunculus amboinicus II 
albus, is apparently very closely allied to, if not identical with, 
Homalomena cordata (Houtt.) Schott, and may be the same as 
Homalomena aromatica (Roxb.) Schott, which is reported from 
Amboina by Engler ; it can scarcely be Homalomena alba Hassk., 
to which Hasskarl reduced it, Neue Schlussel (1866) 128, because 
as far as is known this species is confined to Java. Dracunculus 
amboinicus III ruber should be compared with both Homalomena 
cordata Schott and H. aromatica Schott. Hasskarl, following 
Kunth, reduced it to Homalomena rubescens Kunth, which is 
manifestly an error, as this species is not known from the 
Malay Archipelago. More comprehensive collections are neces- 



ARACEAE 



129 



sary from Amboina before the exact status of these two forms 
can be determined. 

SCH I SMATO GLOTTIS Zollinger and Moritzi 

SCHISMATOGLOTTIS CALYPTRATA (Roxb.) Zoll. & Mor. Syst. Verz. 
(1854) 83. 

Calla calyptrata Roxb. Hort. Beng. (1814) 65, nomen nudum, Fl. Ind. 

ed. 2, 3 (1832) 514. 
Homalomena calyptratum Kunth. Enum. 3 (1841) 57. 
Schismato glottis longipes Miq. Fl. Ind. Bat. 3 (1859) 214. 
Arisarum esculentum Rumph. Herb. Amb. 5: 321, t. Ill, /. 1. 
Amboina, Halong, Robinson PI. Rumph. Amb. 112, September 26, 1913, 
river banks at an altitude of 40 meters, locally known as kasisi. 

Linnaeus originally reduced Arum esculentum, through error, 
to Arum peregrinum Linn., in Stickman Herb. Amb. (1754) 
22, Amoen. Acad. 4 (1759) 131. The type of Calla calyptrata 
Roxb. was from Amboina, and Roxburgh cites the Rumphian 
figure and description in the original description as representing 
his species. Other names involved in the reduction are Colocasia 
humilis Hassk. and Schismato glottis longipes Miq. 

AGLAONEMA Schott 

AGLAONEMA OBLONG I FOLIUM (Roxb.) Kunth Enum. 3 (1841) 55. 

Calla oblongifolia Roxb. Hort. Beng. (1814) 65 (type!); Fl. Ind. ed. 

2, 3 (1832) 516. 
Aglaonema marantifolium Blume Rumphia 1 (1835) 153. 
Scindapsus erectus Presl Epim. (1851) 241 (type!). 
Arum aquaticum Rumph. Herb. Amb. 5: 312, t. 108. 
Adpendix erecta Rumph. Herb. Amb. 5: 487, t. 182, f. 2. 

Not represented in our Amboina collections. The figures and 
the descriptions of both the Rumphian species cited above agree 
closely with Aglaonema oblongif olium Kunth (A. marantifolium 
Blume) as currently interpreted. Arum aquaticum Rumph. is 
the whole basis of Calla oblongifolia Roxb. as originally published 
by Roxburgh, Hort. Beng. (1814) 65, by citation of the Rumphian 
figure; see C. B. Robinson in Philip. Journ. Sci. 7 (1912) Bot. 
414, 419. The description subsequently published by Roxburgh 
was based on specimens cultivated at Calcutta, which originated 
in the Moluccas. Adpendix erecta Rumph. is the basis of Scin- 
dapsus erectus Presl and was also cited by Blume in the original 
description of Aglaonema marantifolium Blume. Linnaeus first 
reduced Arum aquaticum Rumph. to Arum ovatum Linn., in 
Stickman Herb. Amb. (1754) 22, Amoen. Acad. 4 (1759) 131, 
Syst. ed. 10 (1759) 1251, Sp. PL ed. 2 (1763) 1371, but this 
disposition of it was entirely wrong. 

144971 9 



130 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ALOCASIA Necker 

ALOCASIA M ACRORRH IZA (Linn.) Schott Melet. (1832) 18. 

Arum macrorrhizon Linn. Sp. PL (1753) 965. 

Arum mucronatum Lam. Encycl. 3 (1789) 12 (type!). 

Arum silvestre I latifolium Rumph. Herb. Amb. 5: 310. 

Arum indicum sativum Rumph. Herb. Amb. 5: 308, t. 106. 
Amboina, Kati-kati and Halong, Robinson PI. Rumph. Amb. 110, Sep- 
tember, 1913, near streams, altitude 40 to 70 meters, locally known as 
bira and bira puti. 

Arum indicum sativum Rumph. was originally reduced by 
Linnaeus, with doubt, to Arum arborescens Linn., in Stickman 
Herb. Amb. (1754) 22, Amoen. Acad. 4 (1759) 131, but is 
entirely different from this species. Lamarck, Encycl. 3 (1789) 
12, made it the type of his Arum mucronatum. Loureiro, Fl. 
Cochinch. (1790) 536, referred it to his Arum indicum, which 
is supposed to be a synonym of Colocasia indica Engl. Forster, 
PL Esculent. (1768) 58, correctly reduced it to Arum macrorrhi- 
zon Linn. = Alocasia macrorrhiza (Linn.) Schott. Rumphius 
very briefly described three forms; namely, I nigrum, II fuscum, 
and III album, the first of which Hasskarl, Neue Schliissel (1866) 
126, referred to typical Alocasia indica Schott, and the last 
two he referred to Colocasia indica Kunth var. atroviridis Hassk. 
and C. indica Kunth var. pallida Hassk., respectively, but Colo- 
casia indica Kunth, non Engl., is supposed to be a synonym of 
Alocasia indica Schott. The descriptions are so very short that 
it is impossible definitely to determine just what forms were 
intended, but in all probability they were merely variants of 
Alocasia macrorrhiza (Linn.) Schott. I can see no valid reason 
for considering that Arum silvestre I latifolium Rumph., Herb. Amb. 
5: 310 (non t. 107), represents other than a form of Alocasia 
macrorrhiza Schott; see Alocasia longiloba Miq. 

ALOCASIA LONGILOBA Miq. Fl. Ind. Bat. 3 (1857) 207? 

Arum silvestre II medium Rumph. Herb. Amb. 5: 310, t. 107. 
Amboina, Soja, in forests, altitude about 400 meters, Robinson PI. 
Rumph. Amb. 109, August 2, 1913. 

The specimen lacks the spathe, but presents the f ruiting spadix. 
It unquestionably represents the form figured by Rumphius under 
Arum silvestre; and I consider that the figure conforms to Arum 
silvestre II medium rather than to Arum silvestre I latifolium; 
although it has very generally been referred to the latter. Arum 
silvestre I latifolium I take to be a form of Alocasia macrorrhiza 
Schott, above. Hasskarl, Neue Schliissel (1866) .126, reduced 
Arum silvestre II medium to Alocasia montana Schott, certainly 



ARACEAE 



131 



an erroneous disposition of it. The figure has been referred to 
various species — by Linnaeus, through error, to Arum sagitti- 
folium Linn., in Stickman Herb. Amb. (1753) 22, Amoen. Acad. 
4 (1759) 131, Syst. ed. 10 (1759) 1251; by Schott, Prodr. (1860) 
144, to Alocasia indica Schott; and by Miquel, Fl. Ind. Bat. 3 
(1857) 207, with doubt to Alocasia longiloba Miq. Alocasia 
indica Schott is a species of uncertain status; and, as generally 
interpreted, the specific name is probably invalid. It is supposed 
to be Arum indicum Roxb., Fl. Ind. ed. 2, 3 (1832) 498, non Arum 
indicum Lour., Fl. Cochinch. (1790) 536, but the specific name 
should go with Loureiro's species, which is supposed to be Colo- 
casia indica. I have not seen Schott's original publication of 
Alocasia indica. At any rate, the plant figured by Rumphius 
as Arum silvestre has little in common with the one described by 
Roxburgh as Arum indicum or with the one described by Loureiro 
under the same name. 

COLOCASIA Schott 

COLOCASIA ESCULENTA (Linn.) Schott Melet. 1 (1832) 18. 
Arum esculentum Linn. Sp. PI. (1753) 965. 
Arum colocasia Linn. Sp. PI. (1753) 965. 
Colocasia antiquorum Schott Melet. 1 (1832) 18. 
Arum aegyptium Rumph. Herb. Amb. 5: 313, t. 109. 
Caladium aquatile Rumph. Herb. Amb. 5: 318, t. 110, f. 1. 

Taro is not represented in our Amboina collections, although 
doubtless the plant, in several forms, is still cultivated in Am- 
boina as in all parts of the Indo-Malayan region. Like many 
widely cultivated plants, the species is enormously variable ; and, 
being poorly represented in herbaria, no satisfactory arrange- 
ment of the numerous forms and varieties has been proposed, nor 
is any attempted arrangement of these likely to prove satisfactory 
unless based on a comprehensive collection of living plants. Lin- 
naeus originally reduced Arum aegyptium Rumph. to Arum colo- 
casia Linn., and Caladium aquatile Rumph. to Arum esculentum 
Linn., in Stickman Herb. Amb. (1754) 22, Amoen. Acad. 4 
(1759) 131, Syst. ed. 10 (1759) 1251, and by later authors both 
have been referred to the above names and to numerous other 
synonyms of the species. What is apparently the oldest valid 
specific name for the collective species is here adopted. Various 
names involved in the reduction of the forms figured and de- 
scribed by Rumphius are Arum peltatum Lam., Colocasia vera 
Hassk., various proposed varieties of Colocasia antiquorum 
Schott, Caladium esculentum Vent., Caladium nymphaei folium 
Willd., and Caladium esculentum var. aquatilis Hassk. Under 



132 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Arum aegyptium Rumphius describes eight forms or varieties, 
and under Caladium aquatile two, all of which are reducible to 
Colocasia esculenta Schott, sensu latiore. I cannot agree with 
W. F. Wight, Contr. U. S. Nat. Herb. 9 (1905) 206, in interpret- 
ing the genus Caladium as typified by the Rumphian description. 
The type of Caladium, the plant figured, and the one to which the 
description applies, is Caladium bicolor Vent. 

TYPHONIUM Schott 

TYPHONIUM DIVARICATUM (Linn.) Decne. in Nuov. Ann. Mus. Paris 
3 (1834) 367; Blume Rumphia 1 (1835) 130 var. ROBUSTUM 
Kunth. Enum. 3 (1841) 26. 
Arum divaricatum Linn. Sp. PI. ed. 2 (1863) 1369. 
Typhonium javanicum Miq. Fl. Ind. Bat. 3 (1857) 193. 
Arisarum amboinicum Rumph. Herb. Amb. 5: 319, t. 110, f. 2. 

This species is not represented in our Amboina collections. 
The reduction and synonymy follow Engler, in DC. Monog. Phan. 
2 (1879) 612. Linnaeus referred it to Arum trilobum Linn., 
in Stickman Herb. Amb. (1754) 22, Amoen. Acad. 4 (1759) 
131, and to A. trilobatum Linn., in Syst. ed. 10 (1759) 1251, Sp. 
PI. ed. 2 (1763) 1369, in which he was followed by all authors 
until Blume placed it under Typhonium divaricatum Decne. 
Hasskarl, Neue Schlussel (1866) 127, placed Arisarum amboini- 
cum as the equivalent of Typhonium divaricatum Decne.; but, 
probably through oversight, he did not cite the figure. 

PI ST I A Linnaeus 
PISTIA STRATIOTES Linn. Sp. PI. (1753) 963. 
Zala asiatica Lour. Fl. Cochinch. (1790) 405. 
Pistia minor Blume Rumphia 1 (1835) 78 (type!). 
Plantago aquatica Rumph. Herb. Amb. 6: 177, t. 7h, f. 2. 
Plantago aquatica II minor Rumph. Herb. Amb. 6: 177. 

This common and well-known species is not represented in our 
Amboina collections. The reduction to Pistia stratiotes Linn, 
was first made by Linnaeus, in Stickman Herb. Amb. (1754) 28, 
Amoen. Acad. 4 (1759) 136, Syst. ed. 10 (1759) 1249, which has 
been followed by practically all authors. Loureiro, Fl. Cochinch. 
(1790) 405, cites it under Zala asiatica Lour., a synonym of Pistia 
stratiotes Linn. The form described by Rumphius as Plantago 
aquatica II minor is the whole basis of Pistia minor Blume, which 
is manifestly only a reduced form of Pistia stratiotes Linn., due 
to habitat. The plant is enormously variable in size, depending 
on the age of the individuals, habitat, etc., the tendency, where 
the plants are very numerous and crowded, being to a great 
reduction in size. 



LEMNACEAE— BROMELIACEAE 



133 



ARACEAE OF UNCERTAIN STATUS 

Two forms very briefly described or mentioned by Rumphius are entirely 
indeterminable; these are Erva de Sta Maria Rumph., Herb. Amb. 5: 326, 
and Itelpou Rumph., 1. c. 327. Hasskarl thought that the former might be 
a representative of the Araceae, which is probably correct, and that the 
latter might be Brachyspatha variabilis Schott — Amorphophallus variabilis 
Blume ; neither was from Amboina. 

LEMNACEAE 

LEMNA Linnaeus 

LEMNA sp. 

Lens palustris Rumph. Herb. Amb. 6: 178. 

Hasskarl, Neue Schliissel (1866) 180, placed this under Lemna 
minor Linn., following Miquel. Rumphius gives no description, 
and the plant he named may have been any of the Malayan 
Lemnaceae of the genus Lemna or the genus Spirodela. 

FLAGELLARIACEAE 

FLAGELLARIA Linnaeus 

FLAGELLARIA INDICA Linn. Sp. PI. (1753) 333. 

Palmijuncus laevis Rumph. Herb. Amb. 5: 120, t. 59, f. 1. 
Amboina, Paso and Eri, Robinson PL Rumph. Amb. 211, September and 
October, 1913, in thickets near the seashore. 

The original reduction of Palmijuncus laevis Rumph. to Fla- 
gellaria indica was made by Linnaeus, in Stickman Herb. Amb. 
(1753) 20, Amoen. Acad. 4 (1759) 129, Syst. ed. 10 (1759) 989, 
Sp. PI. ed. 2 (1762) 475, which has been followed by all authors, 
and which is certainly the correct disposition of it. 

BROMELIACEAE 

ANANAS Tournefort 

ANANAS COM OS US (Linn.) comb. nov. 

Bromelia ananas Linn. Sp. PI. (1753) 285. 

Bromelia comosa Linn, in Stickman Herb. Amb. (1754) 21, Amoen. 

Acad. 4 (1759) 130 (type!). 
Ananas sativus Schultes f. Syst. 7 (1830) 1283. 
Ananassa sativa Lindl. Bot. Reg. (1827) sub t. 1068. 
Anassa domestica Rumph. Herb. Amb. 5: 227, t. 81. 

Amboina, Koesoekoesoe sereh, Robinson PI. Rumph. Amb. 382, October 3, 
1913, semicultivated, locally known as nanas. 

The common pineapple, as described and figured by Rumphius, 
is the whole basis of Bromelia comosa Linn, as originally pub- 



134 



RUMPHIUS'S HERBARIUM AMBOINENSE 



lished by Linnaeus in the year 1754. It is thus the oldest 
valid specific name for the species. The Rumphian figure, which 
is excellent, has very generally been cited by various authors 
under one or the other of the synonyms given above. The forms 
indicated by Rumphius as mas, femina, and alba are apparently 
merely cultural forms of the common pineapple. 

COMMELINACEAE 

COM M ELI N A Linnaeus 

COM M ELI N A N U Dl FLORA Linn. Sp. PL (1753) 41. 

Arundinella I minor Rumph. Herb. Amb. 6: 23, t. 9, f. 2. 

Amboina, Soja and Roemah tiga, Robinson PI. Rumph. Amb. 4.07, July 
and August, 1913, locally known as kangkong. 

The Rumphian plant has generally been reduced to Commelina 
benghalensis Linn., and the description, in part, seems to apply 
to that species. The figure is very poor, but the description 
is at least sufficiently definite to place the plant Rumphius in- 
tended in Commelina. Burman f., Fl. Ind. (1768) 16, referred 
it to Commelina benghalensis Linn., while Commelina rumphii 
Kostel., Allg. Med.-Pharm. Fl. 1 (1831) 126, may have been 
based wholly or in part on Rumphius (publication not seen) . 

COMMELINA BENGHALENSIS Linn. Sp. PI. (1753) 41. 
Arundinella II major Rumph. Herb. Amb. 6: 24. 
Amboina, near the town of Amboina, Robinson PI. Rumph. Amb. 406, 
July 22, 1913, locally known as kangkong ayer. 

Hasskarl, Neue Schlussel (1866) 155, has suggested Commelina 
communis Linn, as the proper disposition of Arundinella II 
major, but the description appears to me better to apply to 
Commelina benghalensis Linn. 

AN EILEM A R. Brown 

AN El LEM A VITIENSE Seem. var. PETIOLATA C. B. Clarke in DC. 
Monog. Phan. 3 (1881) 220. 
Arundinella IV Rumph. Herb. Amb. 6: 25. 
Amboina, Soja and Kaju poeti, Robinson PI. Rumph. Amb. 408, August, 
1913, along roadsides and in forests. 

Whether the plant is the above form or the very closely 
allied Aneilema monadelphum Kunth, it is undoubtedly the form 
that Rumphius described. Hasskarl, Neue Schlussel (1866) 155, 
expressed the opinion that it was some species of Gramineae, but 
the characters indicated by Rumphius for Arundinella IV all 
apply to this species of Aneilema. 



PONTEDERIACEAE— STEMONACEAE 



135 



CYANOTIS Don 

CYAN OTIS MOLUCCANA (Roxb.) Merr. in Philip. Journ. Sci. 2 (1907) 
Bot. 266. 

Commelina moluccana Roxb. Fl. Ind. ed. 2, 1 (1832) 172. 
Commelina uniflora Hassk. Commel. Ind. (1870) 104. 
Arundinella III aquatica Rumph. Herb. Amb. 6: 24? 

This reduction of Arundinella III aquatica is merely a sugges- 
tion. However, it may be merely Commelina obtusifolia Vahl, 
as suggested by Hasskarl, Neue Schlussel (1866) 155, which 
is supposed to be a synonym of Commelina nudiflora Linn. 

FLOSCOPA Loureiro 

FLOSCOPA SCAN DENS Lour. Fl. Cochinch. (1790) 193. 
Arundinella V Rumph. Herb. Amb. 6: 25. 
Amboina, in sago swamps near the town of Amboina, Robinson PL 
Rumph. Amb. 409, July, 1913, locally known as kangkong ayer. 

The reduction follows Hasskarl, Neue Schlussel (1866) 165, 
who suggests Floscopa paniculata Hassk. (=F. scandens Lour.), 
as the proper disposition of Arundinella V of Rumphius. 

PONTEDERIACEAE 

MONOCHORIA Presl 

MONOCHORIA VAGINALIS (Burm. f.) Presl Rel. Haenk. 1 (1827) 128. 

Pontederia vaginalis Burm. f. Fl. Ind. (1768) 80. 

Olus palustre Rumph. Herb. Amb. 6: 178, t. 75, f. 1. 
Amboina, in grassy pools near the town of Amboina, Robinson PI. 
Rumph. Amb. 225, July 25, 1913. 

Olus palustre was first reduced to Monochoria (Pontederia) 
vaginalis by Linnaeus, Mant. 2 (1771) 222, this certainly being 
the correct disposition of it. 

Olus palustre femina Rumph., 1. c. 178, is probably merely a form 
of Monochoria vaginalis Presl, approaching M. vaginalis var. 
plantaginea Solms, and possibly represented by PI. Rumph. Amb. 
22U from Amboina, August 23, 1913. Hasskarl, Neue Schlussel 
(1866) 180, suggests that Olus palustre femina is Monochoria 
sagittata Kunth=M. hastata Presl, but the description hardly 
warrants this disposition of it. 

STEMONACEAE 

STEM ON A Loureiro 
STEM ON A TUBEROSA Lour. Fl. Cochinch. (1790) 404. 

Roxburghia gloriosoides Roxb. PI. Coromand. 1 (1795) 29, t. 82. 
Ubium polypoides I album Rumph. Herb. Amb. 5: 364, t. 129. 
Amboina, Paso, Robinson PI. Rumph. Amb. 295, October 29, 1913, climb- 
ing over trees along the seashore. 



136 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Ubium polypoides Rumph. was correctly reduced to Stemona 
tuberosa by Loureiro in the original description of that species, 
in which disposition of it later authors have generally concurred, 
including C. H. Wright in his paper on the genus Stemona, 
Journ. Linn. Soc. Bot. 22 (1896) 490-496. 

STEMONA MOLUCCANA (Blume) C. H. Wright in Journ. Linn. Soc. Bot. 
22 (1896) 494. 

Roxburghia moluccana Blume Enum. PI. Jav. 1 (1827) 9 (type!). 
Ubium polypoides II nigrum Rumph. Herb. Amb. 5: 365. 

This species is figured and fully described by Dr. J. J. Smith, 
Ic. Bogor. 3 (1897) 111-114, t 21*5, 2J>6. Wright's and Blume's 
descriptions are entirely inadequate. It is by no means certain 
that the two forms described by Rumphius are really distinct, 
or that the form figured by him, which I have placed under 
Stemona tuberosa Lour., really belongs with Loureiro's species. 
Stemona tuberosa Lour, and S. moluccana C. H. Wright are very 
closely allied. In this connection it is of interest to note that 
Dr. Robinson considered that his specimen represented Ubium 
polypoides II nigrum Rumph. rather than U. polypoides I album 
Rumph. where I have placed it. 

LILIACEAE 

ALOE Linnaeus 
ALOE VERA Linn. Sp. PI. (1753) 320. 

Sempervivium indicum majus Rumph. Herb. Amb. 5: 271. 

This was apparently correctly placed by Henschel, Vita Rumph. 
(1833) 177, who considered it to be Aloe perfoliata Linn. var. 
vera Linn. The species is widely cultivated for medicinal pur- 
poses in the Malayan region. 

D I AN ELLA Lamarck 

Dl AN ELLA ODORATA Blume Enum. 1 (1827) 13. 

Gladiolus odoratus indicus Rumph. Herb. Amb. 5: 185, t. 73. 
Amboina, Way tommo and Soja road, Robinson PI. Rumph. Amb. 505, 
August, 1913, on grassy slopes and barren hills, altitude 50 to 300 meters. 

This was reduced by Linnaeus to Dracaena ensifolia Linn, in 
the original description of that species, Mant. 1 (1767) 63, but 
his description seems to have been based on actual specimens. 
At any rate Dianella ensifolia (Linn.) DC, as currently 
interpreted, is not the same as the form that Rumphius de- 
scribed and figured. Lamarck, Encycl. 2 (1786) 276, cites it 
under Dianella nemorosa, but Dianella nemorosa Lam. was based 
on specimens from Bourbon and Isle of France and is a distinct 
species. Gladiolus odoratus indicus Rumph. is unquestionably 



LILIACEAE 



137 



identical with Dianella odorata Blume, to which it was referred 
by Blume, Schultes, Kunth, and other authors ; see Hallier f ., in 
Nova Guinea 8 (1914) 996. 

PLEOMELE Salisbury 

PLEOMELE ANGUSTI FOLIA (Roxb.) N. E. Br. in Kew Bull. (1914) 277. 
Dracaena angustifolia Roxb. Hort. Beng. (1814) 24 (type!), Fl. Ind. 

ed. 2, 2 (1832) 155. 
Terminalis angustifolia Rumph. Herb. Amb. 4: 81, t. 35. 
Amboina, Amahoesoe, Robinson PI. Rumph. Amb. 506, August 28, 1913, 
near the seashore, locally known as chamara. 

Terminalis angustifolia Rumph. was originally discussed by 
Lamarck, Encycl. 2 (1786) 324, under Dracaena reflexa Lam., 
as possibly representing that species. Lamarck's species, how- 
ever, is quite different from Roxburgh's and is Pleomele reflexa 
N. E. Br. Terminalis angustifolia Rumph. is the whole basis 
of Dracaena angustifolia Roxb. as originally published, in the 
Hortus Bengalensis (1814) 24, by citation of the Rumphian 
figure. The description, later published by Roxburgh, based 
on an Amboina specimen cultivated in the botanic garden at 
Calcutta, unquestionably applies to the same species. Cordyline 
rumphii Hook, is also referable here as a synonym, at least in 
part. 

TAETSIA Medicus 

(Cordyline auct., non Adanson) 

TAETSIA FRUTICOSA (Linn.) comb. nov. 

Convallaria fruticosa Linn, in Stickman Herb. Amb. (1754) 16, 

Amoen. Acad. 4 (1759) 126 (type!), Syst. ed. 10 (1759) 984. 
Asparagus terminalis Linn. Sp. PI. ed. 2 (1762) 450. 
Dracaena terminalis Rich, in Lam. Encycl. 2 (1786) 324. 
Calodracon terminalis Planch. Fl. des Serres I 6 (1850-51) 137. 
Terminalia fruticosa Goepp. in Nov. Act. Acad. Nat. Cur. 25 (1855) 

53. 

Cordyline terminalis Kunth in Abh. Acad. Berlin (1820) 30, Enum. 
5 (1850) 25. 

Taetsia terminalis W. F. Wight in Contr. U. S. Nat. Herb. 9 (1905) 
382. 

Terminalis alba domestica Rumph. Herb. Amb. 4: 79, t. S4, f. 1. 

Terminalis alba silvestris Rumph. Herb. Amb. 4: 80. 

Terminalis rubra Rumph. Herb. Amb. 4: 80, t. 3U, f. 2. 

Terminalis rubra silvestris Rumph. Herb. Amb. 7: 40, t. 20. 
Amboina, Batoe merah, Hoenoet, and Mahija, Robinson PL Rumph. 
Amb. 507, August and October, 1913, on hills and in light forests, altitude 
30 to 150 meters, locally known as pandusti, pandusti puti, and daun 
pandusti. H 

The form cited above is exactly Terminalia alba silvestris 



138 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Rumph., but there is little or no reason for considering that the 
other three forms named by Rumphius represent distinct species. 
Taetsia fruticosa is widely distributed in the Indo-Malayan region 
and presents considerable variation in its cultivated forms, 
especially in the color of its leaves, which vary from green to 
reddish or purple. "Terminalis Rumph. amb. 4 p. 79. t. 34" 
is the whole basis of Convallaria fruticosa Linn. (1754), but the 
original description of Asparagus terminalis Linn, was mani- 
festly based on an actual specimen; Terminalis Rumph. is cited 
as a synonym. On plate 3k f two forms are figured by Rumphius, 
fig. 1 representing the plant with greenish leaves (T. alba 
domestica), and fig. 2, the plant with reddish or purplish 
leaves (T. rubra) ; the latter is manifestly only a color variant 
of the former. Kunth, Enum. 5 (1850) 25, considered that Ter- 
minalis rubra did not belong to Cordyline (Taetsia) or even 
in the Liliaceae, but the figure certainly represents the common 
and well-known Cordyline terminalis Kunth = Taetsia fructicosa 
(Linn.) Merr. I consider Terminalis alba silvestris Rumph. 
to be referable to the same species as the other forms described 
by Rumphius. The chief objection to the reduction of Termi- 
nalis rubra silvestris to Taetsia fruticosa is that the figure re- 
presents the nerves of the leaves as altogether too prominent; 
but this may be due to an error on the part of the artist. 
Hasskarl, Neue Schlussel (1866) 190, suggested that it was the 
same as Cordyline jacquinii Kunth var. rubens* Hassk., but 
Cordyline jacquinii Kunth is supposed to be a synonym of C. 
terminalis. 

Taetsia Medic, is here deliberately accepted in place of Cordy- 
line as the proper generic name of this plant, as it is manifestly 
the oldest valid one for the genus, in spite of the fact that 
Cordyline Commers., ex Juss. Gen. (1789) 41, is retained in the 
list of nomina conservanda adopted by the Vienna Botanical 
Congress in preference to Terminalis Rumph. The Rumphian 
designation has no standing as a generic name. Cordyline 
Adanson (1763) is the same as Sansevieria Thunb. and ante- 
dates Thunberg's name. In adopting Taetsia, I agree with N. 
E. Brown,* who, while retaining Cordyline, states: "At the 
same time, however, my personal view of the case would be 
to abolish the use of the name Cordyline altogether, in con- 
sequence of the great confusion connected with it, and replace 
it by that of Taetsia." 



* Kew Bull. (1914) 275. 



LILIACEAE 



139 



SMI LAX Linnaeus 

SMI LAX JAVENSIS A. DC. in DC. Monog. Phan. 1 (1878) 175. 
Pseudochina amboinensis Rumph. Herb. Amb. 5: 437, t. 161. 
Amboina, Amahoesoe, Robinson PI. Rumph. Amb. 50U, September 18, 
1913, in thickets along the seashore, ascending to an altitude of at least 
40 meters, locally known as tali baduri. 

This was originally reduced by Linnaeus to Smilax china Linn,, 
in Stickman Herb. Amb. (1754) 24, but in Amoen. Acad. 4 
(1759) 133, the reference to Smilax china is excluded, although 
Ubium nummular ium, t. 162 {=Dioscorea nummularia Lam.!), 
is erroneously given as the equivalent of Smilax china Linn. In 
his Mantissa 2 (1771) 499, Linnaeus again erroneously referred 
Pseudochina amboinensis Rumph. to Smilax zeylanica Linn., 
which was followed by all subsequent authors up to the year 
1878, when A. de Candolle suggested that it might be the same as 
Smilax javensis A. DC. I know A. de Candolle's species only by 
description, but the description applies very closely to the Am- 
boina specimen cited above, so that this is probably the correct 
disposition of Pseudochina amboinensis Rumph. However, both 
Smilax javensis A. DC. and the Amboina specimen should be 
critically compared with the Australian Smilax australis R. 
Br., which is a very closely allied and, perhaps, identical form. 

SMILAX LEUCOPHYLLA Blume Enum. PL Jav. 1 (1827) 18. 
Pseudochina nigra Rumph. Herb. Amb. 5: 439. 
Amboina, Negri lama, Robinson PI. Rumph. Amb. 503, August 8, 1913, 
climbing on trees, altitude about 15 meters, locally known as tali baduri. 

This appears to be the correct disposition of Pseudochina 
nigra Rumph., although Hasskarl, Neue Schlussel (1866) 144, 
considered it to be referable to the Australian S. glycyphylla 
Sm., a species known only from Australia, and quite different 
from S. leucophylla Blume. The Philippine Smilax vicaria 
Kunth is probably not specifically distinct, from Blume's species. 

SMILAX LEUCOPHYLLA Blume var. PLATYPHYLLA var. nov. 
Pseudochina alba latifolia Rumph. Herb. Amb. 4: 438? 
Amboina, Lateri, Robinson PI. Rumph. Amb. 502 (type), September 9, 
1913, in forests, altitude about 250 meters, locally known as tali baduri. 

Ramis distanter crasse aculeatis; foliis coriaceis, usque ad 
30 cm longis et 16 cm latis. 

This may prove to be specifically distinct from Smilax leuco- 
phylla Blume when more material is available for study. It 
is well characterized by its very large leaves, long petioles, and 



140 RUMPHIUS'S HERBARIUM AMBOINENSE 

long infructescences. The petioles, including the very promi- 
nent sheathing base, are 5 to 6 cm in length. The infructes- 
cences are about 20 cm long, with at least four long-peduncled 
umbels. Fruits globose, about 1 cm in diameter. 

It may or may not be the same as Pseudochina alba latifolia 
Rumph., but it is quite certain that the Rumphian plant is not 
the same as Smilax villandia Ham. =S. indica Vitm. under which 
name it is briefly discussed by Hasskarl. 

SMILAX CHINA Linn. Sp. PI. (1753) 1029? 

Smilax sarmentis spinulosis etc. Rumph. Herb. Amb. 7: 72, t. 30? 
Radix chinae Rumph. Herb. Amb. 5: 441? 

The identity of the two Rumphian synonyms cited above can 
only be surmised. Hasskarl, Neue Schliissel (1866) 144, con- 
siders Radix chinae to be the same as Smilax aspera Linn., and 
the other as possibly representing S. bauhinioides Kunth. Plate 
30 of the Auctuarium is missing in our copy of the Herbarium 
Amboinense. 

AMARYLLIDACEAE 
CRINUM Linnaeus 

CRINUM ASIATICUM Linn. Sp. PI. (1753) 292. 

Crinum toxicarium Roxb. Hort. Beng. (1814) 23 (type!), Fl. Ind. 

ed. 2, 2 (1832) 134. 
Radix toxicaria I major Rumph. Herb. Amb. 6: 155, t. 69. 

Amboina, Amahoesoe, Robinson PL Rumph. Amb. 130, August 30, 1913, 
along the seashore, locally known as bauang laut. 

Radix toxicaria Rumph. was originally reduced by Linnaeus 
to Crinum asiaticum Linn., in Stickman Herb. Amb. (1754) 28, 
Amoen. Acad. 4 (1759) 136, Syst. ed. 10 (1759) 976, a reduction 
that is certainly correct, and one that has been accepted by 
most authors. It is the type of Crinum toxicarium Roxb., which 
was based wholly on the Rumphian figure and description, Hort. 
Bengal. (1814) 23; see C. B. Robinson in Philip. Journ. Sci. 7 
(1912) Bot. 413. It was cited later in the very brief description 
given in the second edition of the Flora Indica. The species 
is very widely distributed along the seashores of the Indo- 
Malayan and Polynesian regions; it is exceedingly variable in 
size, depending on the age of the plant and on its habitat. The 
form very briefly described by Rumphius as Radix toxicaria III 
montana, 1. c. 156, from Ceram, is probably merely a dwarfed 
form of Crinum asiaticum Linn. 



AMARYLLIDACEAE 



141 



CRINUM RUMPHII sp. nov. 

Radix toxicaria II terrestris Rumph. Herb. Amb. 6: 156. 

Amboina, Hitoe lama, Robinson PL Rumph. Amb. 131 (type) , October 
8, 1913, in forests, altitude about 200 meters, locally known as pohon tolok. 

Planta magna, glabra; foliis ut videtur numerosis, usque ad 
70 cm longis et 18 cm latis, chartaceis, acutis, basi angustatis, 
petiolo circiter 30 cm longo. Floribus numerosis, breviter pe- 
dicellatis, tubo circiter 15 cm longo, segmentis lineari-lanceolatis, 
circiter 14 cm longis et 6 mm latis. 

A large, entirely glabrous plant. Leaves apparently numer- 
ous, when dry dark-olivaceous chartaceous, about 70 cm long 
and 18 cm wide, acute, base narrowed, the petiole about 20 cm 
long, and when dry and flattened out 2.5 to 3 cm wide. Peduncle 
not seen. Flowers numerous, white, at least 20 to each peduncle, 
the spathe-valves about 18 cm long and 3 cm wide, narrowed 
upward, subacute. Pedicels 1 to 1.5 cm long, the perianth-tube 
slender, including the ovary about 15 cm long. Flowers white, 
the filaments lilac. Segments linear-lanceolate, about 14 cm 
long and 6 mm wide, acute. Fruit not seen. 

A species well characterized by its very large leaves and long 
slender perianth-tube. It is manifestly in the same group as 
Crinum asiaticum Linn., but differs from that species in many 
characters, as well as in its entirely different habitat. Hasskarl, 
Neue Schliissel (1866) 178, thought that Radix toxicaria II 
terrestris might be the same as Crinum procerum Carey, which, 
however, is a synonym of C. asiaticum Linn. 

CRINUM ZEYLANICUM Linn. Syst. ed. 12 (1767) 236. 
Amaryllis zeylanica Linn. Sp. PI. (1753) 293. 
Amaryllis lineata Lam. Encycl. 1 (1783) 123. 
Tulipa javana Rumph. Herb. Amb. 5: 306, t. 105.' 

This species is not represented in our Amboina collections, 
although doubtless it is still cultivated in Amboina as it is in 
other parts of Malaya. Rumphius states that it was introduced 
into Amboina from Java about 1670. The figure is an excellent 
representation of Crinum zeylanicum Linn. It was first reduced 
by Linnaeus to Amaryllis zeylanica Linn., in Stickman Herb. 
Amb. (1754) 22, Amoen. Acad. 4 (1759) 131, Syst. ed. 10 (1759) 
977, Sp. PL ed. 2 (1762) 421, which as Crinum zeylanicum Linn, 
has very generally been accepted as the correct disposition of 
Tulipa javana Rumph. 



142 



RUMPHIUS'S HERBARIUM AMBOINENSE 



EURYCLES Salisbury 

EURYCLES AM BOI N ENSIS (Linn.) Lindl. in Loud. Encycl. PI. (1829) 

242. 

Pancratium amboinense Linn. Sp. PI. (1753) 291. 

Pancratium narbonense Linn, in Stickman Herb. Amb. (1754) 28 
(type!). 

Amaryllis rotundifolia Lam. Encycl. 1 (1783) 124. 

Eurycles silvestris Salisb. in Trans. Hort. Soc. 1 (1812) 337. 

Cepa silvestris Rumph. Herb. Amb. 6: 160, t. 70, f. 1. 

This widely distributed, well-known, characteristic species is 
not represented in our Amboina collections. Linnaeus, in Stick- 
man Herb. Amb. (1754) 28, first reduced Cepa silvestris Rumph. 
to Pancratium narbonense Linn. ; this name, apparently a lapsus 
calami for P. amboinense, is not listed in Index Kewensis and, 
being based wholly on Rumphius, thus becomes a synonym of 
Eurycles amboinensis Lindl. In the reprint of Stickman's 
paper, Amoen. Acad. 4 (1759) 136, Syst. ed. 10 (1759) 976, Sp. 
PL ed. 2 (1762) 419, it is properly placed under Pancratium 
amboinense lAxm=Eurycles amboinensis Lindl. Willdenow, 
Sp. PL 2 (1799) 47, cites it under Crinum nervosum L'Herit. ; 
and other authors cite it under Amaryllis rotundifolia Lam., 
Eurycles coronata Salisb., E. nervosa Roem., and E. silvestris 
Salisb. — all synonyms of E. amboinensis (Linn.) Lindl. 

CURCULIGO Gaertner 
CURCULIGO ORCHOIDES Gaertn. Fruct. 1 (1788) 63, t. 16. 
Curculigo rumphiana Schultes Syst. 7 2 (1830) 757 (type!). 
Orchis amboinica major II Rumph. Herb. Amb. 6: 117, t. 5£. f. 1. 

Amboina, Batoe gadjah, Robinson PI. Rumph. Amb. 128, August 5, 1913, 
on open hillsides, altitude about 150 meters. 

Orchis amboinica major II, as described and figured by 
Rumphius, was cited by Gaertner in the original description of 
Curculigo orchoides Gaertn. as a synonym of his species, and 
the Amboina specimen cited above agrees with the species as 
currently interpreted. Most authors have followed Gaertner 
in this reduction, but Schultes,- in a discussion following the 
description of Curculigo orchoides, considers the Rumphian 
plant to represent a distinct species, which he called Curculigo 
rumphiana Schultes and which was based wholly on the Rumph- 
ian reference cited above. Curculigo rumphiana Schultes is 
not listed in Index Kewensis, and I consider it to be merely a 
synonym of the much older name, Curculigo orchoides Gaertn. 

CURCULIGO CAPITULATA (Lour.) O. Kuntze Rev. Gen. PI. 1 (1891) 
703. 

Leucojum capitulatum Lour. Fl. Cochinch. (1790) 199. 



AMAKYLLIDACEAE 



143 



Curculigo recurvata Dryander in Ait. Hort. Kew. ed. 2, 2 (1811) 253. 
Curculigo sumatrana Roxb. Hort. Beng. (1814) 24 (type!). 
Involucrum s. angraecum terrestre tertium Rumph. Herb. Amb. 6: 
114, t. 53. 

Amboina, Kati-kati, Robinson PL Rumph. Amb. 129, October 7, 1913, 
on banks, altitude about 80 meters. 

This was cited by Loureiro, Fl. Cochinch. (1790) 13, under 
Phyllodes placentaria Lo\ir.=Ph r rynium capitatum Willd., a spe- 
cies that is entirely different from Curculigo capitulata 0. 
Kimtze. Poiret, in Lam. Encycl. Suppl. 5 (1817) 645, referred 
it to Curculigo latifolia Dry., in which he was followed by 
numerous authors. Henschel and Pritzel refer it to Curculigo 
recurvata Dry. It is the actual type of Curculigo sumatrana 
Roxb., as published in Hort. Beng. (1814) 24, by citation of tho 
Rumphian figure; see C. B. Robinson in Philip. Journ. Sci. 
7 (1912) Bot. 413. Curculigo sumatrana Roxb., as actually 
described from Sumatran specimens later, Roxburg Fl. Ind. ed. 
2, 2 (1832) 146, seems to be distinct and is generally cited as a 
synonym of Curculigo latifolia Dryand. 

PANCRATIUM Dillenius 

PANCRATIUM ZEYLANICUM Linn. Sp. PI. (1753) 290. 
Lilium indicum Rumph. Herb. Amb. 6: 161, t. 70, /. 2. 

This common and well-known species, excellently figured by 
Rumphius, is not represented in our Amboina collections. It 
occurs here and there in the vicinity of towns and habitations 
throughout the Malayan region. Lilium indicum Rumph. was 
first reduced to Pancratium zeylanicum Linn., in Stickman Herb. 
Amb. (1754) 28, Amoen. Acad. 4 (1759) 136, Sp. PI. ed. 2 
(1762) 417, which is certainly the correct disposition of it, 
and which has been accepted by subsequent authors. 

The form described by Rumphius, in the same chapter, as 
Lilium indicum javanicum, Herb. Amb. 6: 162, a Javan plant with 
yellow flowers, is indeterminable from the data at present avail- 
able. Hasskarl, Neue Schlussel (1866) 178, has suggested that 
it might be Calostemma luteum Sims, but it is entirely im- 
probable that this Australian species had been introduced into 
Java at the time when Rumphius wrote his account. 

POLIANTHES Linnaeus 

POLIANTHES TUBEROSA Linn. Sp. PI. (1753) 316. 
Arnica nocturna Rumph. Herb. Amb. 5: 285, t. 98. 

The tuberose is not represented in our Amboina collections, 
but is probably still cultivated in Amboina, as it is in various 



144 



RUMPHIUS'S HERBARIUM AMBOINENSE 



parts of the Malayan region and the Philippines. Arnica noc- 
turia Rumph. was first reduced to Polianthes tuberosa Linn, 
by Linnaeus, in Stickman Herb. Amb. (1754) 22, Amoen. Acad. 
4 (1759) 131, Syst. ed. 10 (1759) 984, Sp. PI. ed. 2 (1762) 453, 
which is certainly the correct disposition of it, and which has 
been accepted by all subsequent authors. 

AGAVE Linnaeus 

AGAVE CANTALA Roxb. Hort. Beng. (1814) 25, nomen nudum. 
Furcraea cantala Haworth Syn. PI. Succul. Suppl. (1819) 42. 
Agave cantula Roxb. Fl. Ind. ed. 2, 2 (1832) 167. 
Agave rumphii Hassk. in Hoev. & DeVriese Tijdschr. 10 (1843) 121. 
Aloe americana Rumph. Herb. Amb. 5: 273, t. 94. 

The common maguey plant is not represented in our Amboina 
collections. Aloe americana Rumph. was originally reduced by 
Linnaeus to Aloe vivipara Linn., in Stickman Herb. Amb. (1754) 
21, Amoen. Acad. 4 (1759) 130, while in the Syst. ed. 10 (1759) 
986, Sp. PI. ed. 2 (1762) 461, he placed it under Agave vivipara 
Linn, in which he was followed by various authors. This, 
however, seems to be a different species, originally based on 
material from tropical America. Agave cantala Roxb., as 
originally used, is a nomen nudum, but the spelling of the specific 
name is fixed by Haworth's publication of Furcraea cantala 
in 1819, with a reference to Roxburgh. However, Roxburgh 
himself published the name, with a brief description and a 
reference to Rumphius, as Agave cantula Roxb. Agave rumphii 
Hassk. is apparently typified by Aloe americana Rumph. While 
the species is of Mexican origin, having been introduced into the 
Philippines at an early date by the Spaniards, the Indo-Malayan 
form has not as yet been satisfactorily connected with any 
described American species. 

TACCACEAE 

TACCA * Forster 

TACCA PI N N ATI Fl DA Forst. Char. Gen. (1776) 70, t. 85. 
Tacca dubia Schultes Syst. 7 1 (1829) 167 (type!). 
Tacca litorea Rumph. Herb. Amb. 5: 328, t. 

Tacca phailifera Rumph. Herb. Amb. 5: 326 p. p., t. 113, f. 1, a, b. 

This species is not represented in our Amboina collections. 
The illustrations given by Rumphius of Tacca litorea, and of T. 
phailifera t. 113, f. 1, are both excellent representations of the 



* Retained name, Brussels Congress; Leontopetaloides Boehm. (1760) is 
older. 



DIOSCQREACEAE 



145 



common and well-known Tacca pinnatifida Forst. The species 
was confused by Rumphius with Amorphophallus campanulatus 
Blume (see p. 127), Tacca phallifera Rumph. being made up 
of Tacca pinnatifida Forst. with the inflorescence of Amorpho- 
phallus campanulatus Blume, which was described by Rumphius 
as Taccae fungus. Tacca sativa Rumph., I. c. 5 : 324, is also 
apparently a mixture of Tacca pinnatifida Forst. and Amorpho- 
phallus campanulatus Blume, but the description for the most 
part and the figure are Amorphophallus, not Tacca. Tacca 
dubia Schultes was based wholly on Tacca phallifera Rumph,, 
excluding Taccae fungus and t. 113, f. 2, and is manifestly 
nothing but Tacca pinnatifida Forst. Forster apparently took 
his generic name from Rumphius, and in the original publication 
of the species he cites both Tacca sativa Rumph., t. 112, and 
Tacca litorea Rumph., t. 11 U, as synonyms. The type, however, 
was a Polynesian specimen. 

TACCA PALM ATA Blume Enum. PL Jav. 1 (1827) 83. 
Tacca montana Schultes Syst. 7 1 (1829) 168. 

Tacca rumphii Schauer in Nov. Act. Acad. Nat. Cur. 19 (1843) Suppl. 
1: 442. 

Tacca montana Rumph. Herb. Amb. 5: 329, t. 115. 

This species is not represented in our Amboina collections. 
I consider that both forms described by Rumphius — I minor 
and II major — are referable here. The reduction was first made 
by Blume. Schultes merely substituted Rumphius's name for 
that proposed by Blume, reducing Tacca palmata Blume as a 
synonym. Tacca rumphii Schauer was based on Luzon speci- 
mens, manifestly the same as Tacca palmata Blume, with the 
addition of a reference to Tacca montana Rumph. Herb. Amb. 
5: 329, t. 115. 

DIOSCOREACEAE 

DIOSCOREA Linnaeus 

Rumphius described and figured a number of forms and 
species of Dioscorea, under the general name Ubium, which have 
been in part more or less misunderstood by subsequent authors. 
But two species are represented in our Amboina collections. 
These are both common and well-known species, so that the 
material available for study hardly assists in clearing up species 
of doubtful status so far as those based wholly or partly on 
Rumphius's descriptions and figures are concerned.* 

*Prain and Burkill. A synopsis of the Dioscoreas of the Old World, 
Africa excluded, with descriptions of new species and of varieties. Journ. 
As. Soc. Beng. II 10 (1914) 1-41. 

144971 10 



146 



RUMPHIUS'S HERBARIUM AMBOINENSE 



DIOSCOREA BULBIFERA Linn. Sp. PL (1753) 1033. 

Ubium pomiferum Rumph. Herb. Amb. 5: 354, t. 124. 

This species is not represented in our Amboina collections. 
The figure, however, unmistakably represents Dioscorea bulbi- 
fera Linn. The reduction was first made by Linnaeus, in Stick- 
man Herb. Amb. (1754) 23, Amoen. Acad. 4 (1759) 131, Syst. 
ed. 10 (1759) 1294, which has been followed by all authors. 
Ubium pomiferum silvestre Rumph., 1. c. 354, is probably merely a 
form of the same species. Mr. Burkill notes that the wild form 
mentioned by Rumphius is in all probability Dioscorea bulbifera 
Linn. var. vera Prain and Burkill, and that the cultivated forms 
may include var. sativa Prain and Burkill and var. suavior Prain 
and Burkill. 

DIOSCOREA ALATA Linn. Sp. PI. (1753) 1033. 

Ubium vulgare Rumph. Herb. Amb. 5: 346, t. 120. 

Ubium digitatum Rumph. Herb. Amb. 5: 350, t. 121. 

Ubium anguinum Rumph. Herb. Amb. 5: 351, t. 122. 

Ubium ovale Rumph. Herb. Amb. 5: 356, t. 125. 

Ubium draconum Rumph. Herb. Amb. 5: 351, t. 122, f. D, E. 

Ubium anniversarium Rumph. Herb. Amb. 5: 353, t. 123. 

Inhame St. Thome Rumph. Herb. Amb. 5: 355. 

This commonly cultivated yam is not represented in our 
Amboina collections. However, I have little doubt that the 
six forms figured by Rumphius as Ubium vulgare, U. digitatum, 
U. anguinum, U. ovale, TJ. draconum, and U. anniversarium are 
all referable to Dioscorea alata Linn., including the several forms 
described under the first, second, and fourth. The species is 
enormously variable in the shape, color, and size of its sub- 
terranean parts, but is apparently fairly uniform in its vege- 
tative and floral characters. The first three were originally and 
erroneously reduced by Linnaeus to Dioscorea oppositifolia 
Linn., in Stickman Herb. Amb. (1754) 22, 23, Amoen. Acad. 4 
(1759) 131, but later authors have generally cited them as 
synonyms of Dioscorea alata Linn., where they properly belong. 
Ubium ovale Rumph. has been cited by some authors as a 
synonym of Dioscorea bulbifera Linn., but from the figure and 
description it is apparently merely the bulbil-bearing form of 
Dioscorea alata Linn. Ubium anniversarium Rumph. has been 
quoted by Henschel, Hasskarl, Kunth, and Miquel as a possible 
synonym of Dioscorea spiculata Blume, but Mr. I. H. Burkill has 
called my attention to the fact that it is the same form as the 
curious race of Dioscorea alata figured by him in Gard. Bull. 
Straits Settl. 1 (1915) 301, figs. 2-6, (1917) 393, pi 5, 6; and 



DIOSCOREACEAE 



147 



in the Philip. Agr. and Forester 3 (1915) 207, plate 2, figs. 12, 
U, 18. 

In a recent letter to me Mr. Burkill states : 

Of the Philippine Dioscorea alata I have in Singapore in a few races the 
tuberous roots do not respond to geotropism in the normal way, but ascend 
to the surface of the soil where conditions are apt to kill them. If 
one continues to protect them by covering them with earth they continue 
to grow and may become greatly elongated. The drawings of the yams 
on page 30 of Gardens' Bulletin were made at a time when I had not 
discovered how to earth them up. If you will examine the figure of 
Rumphius's Ubium anguinum you will note that what I take to be the same 
race is represented. I believe that this type of yam arose and was propa- 
gated by planting it in the midden at the back door and that the yams 
continued to grow upward with the accumulation of rubbish. 

Mr. Burkill calls my attention to the fact that Inhame St. 
Thome of Piso is Dioscorea alata Linn., and that Rumphius was 
wrong in ascribing it to his Ubium pomiferum. 

DIOSCOREA ESCU LENTA (Lour.) Burkill in Gard. Bull. Straits Settl. 1 
(1917) 396. 

Oncus esculentus Lour. Fl. Cochinch. (1790) 194. 

Dioscorea combilium Ham. in Wall. Cat. (1832) no. 5103 A. 

Dioscorea fasciculata Roxb. Hort. Beng. (1814) 72, nomen nudum, 

Fl. Ind. ed. 2, 2 (1832) 801. 
Dioscorea tiliae folia Kunth Enum. 5 (1840) 401. 
Combilium Rumph. Herb. Amb. 5: 357, t. 126. 

This species is not represented in our Amboina collections. 
The figure, however, unmistakably represents the form commonly 
named Dioscorea tiliaefolia Kunth, but for which Prain and 
Burkill have recently adopted the name Dioscorea aculeata Linn. 
Sir David Prain, however, has discovered that Dioscorea aculeata 
Linn, is the valid name for D. wallichii Hook. f. and that Oncus 
esculentus Lour, supplies the oldest valid name for the species 
under discussion. Linnaeus originally reduced Combilium to 
Dioscorea aculeata Linn., in Stickman Herb. Amb. (1754) 23, 
Amoen. Acad. 4 (1759) 131, Syst. ed. 10 (1759) 1293. Mr. 
Burkill, who has called my attention to the necessary change in 
the specific name of this species, considers that Combilium is 
referable here with the possible exception of Combilium rubrum, 
the status of which is uncertain, and C. tsjampadaha, which is 
described as if a different species. 

DIOSCOREA PENTAPHYLLA Linn. Sp. PI. (1753) 1032. 

Ubium quinquefolium Rumph. Herb. Amb. 5: 359, t. 127. 

This species is not represented in our Amboina collections. 
The figure probably represents the var. malaica Prain & Burkill, 



148 



RUMPHIUS'S HERBARIUM AMBOINENSE 



in Journ. As. Soc. Beng. II 10 (1914) 23. The original reduction 
of Rumphius's plant was made by Linnaeus, in Stickman Herb. 
Amb. (1754) 23, Amoen. Acad. 4 (1759) 131, Syst. ed. 10 (1759) 
1293, and has been accepted by most authors. The forms de- 
scribed as album, rubrum, and fuscum are probably but variants 
of the same species. 

DIOSCOREA HISPIDA Dennst. Schlussel Hort. Malabar. (1818) 33. 
Dioscorea hirsuta Roth Nov. PI. Sp. (1821) 370. 

Dioscorea daemona Roxb. Hort. Beng. (1814) 72, nomen nudum, Fl. 

Ind. ed. 2, 3 (1832) 805. 
Dioscorea triphylla auctt., non Linn. 
Ubium silvestre Rumph. Herb. Amb. 5: 361, t. 128. 
Colot e Philippinis Rumph. 1. c. 364. 
Amboina, Wae, Robinson PI. Rumph. Amb. 375, November 26, 1913, in 
light forests, altitude about 5 meters, locally known as ondo. 

Dioscorea triphylla Linn., Sp. PL (1754) 1032, typified by 
Rheede, Hort. Malabar. 7 : 63, t. 33, is merely a form of Dioscorea 
pentaphylla Linn, and must be considered as a synonym of that 
species. Prain and Burkill, however, propose to cite Dioscorea 
triphylla Linn, as published in Stickman, Herb. Amb. (1754) 23, 
Amoen. Acad. 4 (1759) 131, thus typified by the Rumphian 
figure, and to retain Dioscorea triphylla Linn, as a valid species. 
This usage is inadmissible, and the specific name triphylla Linn, 
should be dropped. The oldest valid specific name appears to be 
Dioscorea hispida Dennst. 

DIOSCOREA NUMM ULARI A Lam. Encycl. 3 (1789) 231 (type!). 

Dioscorea nummular i folia Henschel Vita Rumph. (1833) 183 (type!). 

Dioscorea glabra Koord. Meded. Lands Plantent. 19 (1898) 312. 

Ubium nummularium frugiferum Rumph. Herb. Amb. 5: 444, t. 162. 
Amboina, Negri lama and Soja, Robinson PI. Rumph. Amb. 374, Sep- 
tember 8, 1913, in thickets and on hillsides, altitude 20 to 375 meters, 
locally known as ubi utan. 

Ubium nummularium Rumph. was originally and erroneously 
reduced by Linnaeus to Dioscorea villosa Linn., in Stickman 
Herb. Amb. (1754) 24, Syst. ed. 10 (1759) 1294; but in Amoen. 
Acad. 4 (1759) 133, it was erroneously referred to Smilax china 
Linn, by confusion with the preceding plate. It is the whole 
basis of Dioscorea nummularia Lam., the Rumphian figure and 
description being the type of that species. "Dioscorea num- 
mularifolium Linn." as quoted by Henschel does not exist 
and is a manifest error, the intention being Dioscorea num- 
mularia Lam. The species is well known, being widely dis- 
tributed in the Philippines and the Moluccas and extending to the 
southeast as far as New Guinea and Prince of Wales Island. 



MUSACEAE 



149 



DIOSCOREA sp. aff. D. NUMMULARIA Lam. 

Ubium nummularium floriferum Rumph. Herb. Amb. 5: 445, t. 163. 

This is not represented in our Amboina collections. Miquel, 
Fl. Ind. Bat. 3 (1858) 572, cites this plate with the preceding 
one under Dioscorea nummularia Lam. Hasskarl, Neue Schlus- 
sel (1866) 144, cites the Rumphian name, but not the plate. 
The figure presents a plant with oblong-ovate, prominently acu- 
minate, 5-nerved leaves, which are rounded at the base, in aspect 
somewhat resembling some forms placed under Dioscorea num- 
mularia Lam., but with the stems prominently spiny. I have 
a single Philippine specimen, Wenzel 1120 from Leyte, that 
Burkill has identified as a form of Dioscorea nummularia Lam., 
which has spiny stems, but otherwise all the material of typical 
Dioscorea nummularia Lam. that I have seen is entirely un- 
armed. In aspect the two forms figured by Rumphius are very 
distinct, but the differences may be more apparent than real. 
Possibly the differences in the presence or absence of spines on 
the stems may be due to the fact that the figure of Ubium num- 
mularium frugiferum presents only the upper part of the plant, 
while that of Ubium nummularium floriferum presents also the 
lower part of the plant, the upper part being unarmed. Regard- 
ing this form Mr. Burkill writes as follows : 

Figure 163 is unrecognizable. It may be a very badly drawn Dioscorea 
of the D. nummularia alliance. The type of the inflorescence, the opposite 
leaves, and the characters of the underground parts all permit of this view, 
in which case we need not suppose that the foliage is that of a Stemona, 
but that the artist was very careless in making the drawing. Indeed the 
foliage looks as if it were withered when the drawing was made. 

MUSACEAE 

MUSA Linnaeus 

MUSA PARADISI ACA Linn. Sp. PI. (1753) 1043. 

Musa domestica Rumph. Herb. Amb. 5: 125, t. 60. 

Musa uranoscopos Rumph. Herb. Amb. 5: 137, t. 61, f. 2. 

Musa alphurica Rumph. Herb. Amb. 5: 138, t. 61, f. 3. 

No attempt is here made to refer the numerous forms of the 
banana described by Rumphius to the various named varieties 
and forms of this species. Under Musa domestica Rumphius 
describes sixteen forms, four of which are figured; these for 
the most part are indicated by their native names. Rumphius 
described as distinct "species" Musa uranoscopos and M. alphu- 
rica, the former being reduced by Warburg to subsp. troglody- 
tarum (Linn.) Baker, and the latter to subsp. sapientum (Linn.) 
0. Kuntze; Musa troglodytarum Linn.,' Sp. PI. ed. 2 (1763) 



150 



RUMPHIUS'S HERBARIUM AMBOINENSE 



1478, is typified by Musa uranoscopos Rumph. Herb. Amb. 
5: 137, t. 61, f. 2. From data and material at present available 
no satisfactory arrangement of the various forms of the cul- 
tivated and semicultivated bananas is possible, and the definite 
characters of the numerous species, subspecies, varieties, and 
forms can be determined only by a critical study of ample 
living material. No bananas are represented in our Amboina 
collections. 

MUSA TEXTILIS Nee in Ann. Cienc. Nat. 4 (1801) 123. 

Musa silvestris mindanauensis Rumph. Herb. Amb. 5: 139. 

This was described from Mindanao and Sangir material and 
is manifestly a form of the abaca plant. It is certain that 
Musa textilis Nee includes a number of distinct forms or va- 
rieties, and perhaps a critical study of living material will show 
that some of the forms are worthy of specific rank. 

MUSA TEXTILIS Nee var. AM BOI N ENSIS Warb. in Engl. Pflanzenreich 
1 (1900) 19. 

Musa amboinensis Miq. Fl. Ind. Bat. 3 (1858) 588 (type!). 
Musa silvestris amboinensis Rumph. Herb. Amb. 5: 139. 

The status of this Amboina form is uncertain, and it may 
prove to be specifically distinct from Musa textilis Nee. 

MUSA ACUMINATA Colla in Mem. Acad. Torin. 25 (1820) 338 (Mem. 
Gen. Musa 66). 

Musa rumphiana Kurz in Journ. Agric. Hort. Soc. India II 5 (1878) 
164. 

Musa simiarum Miq. Fl. Ind. Bat. 3 (1858) 589. 

Musa simiarum Rumph. Herb. Amb. 6: 138, t. 61, f. 1. 

This sylvan species, briefly described by Rumphius, is probably 
a valid one, as considered by Warburg, in Engl. Pflanzenreich 
1 (1900) 21, under the name Musa acuminata Colla. It has been 
referred by various authors to Musa seminifera Lour., M. dis- 
color Horan., and M. paradisiaca Linn, var., but these reductions 
are certainly erroneous. 

H ELICON I A * Linnaeus 

(By Th. Valeton) 

H ELICON I A BIHAI Linn. Mant. 2 (1771) 211. 
Musa bihai Linn. Sp. PI. (1753) 1043. 

Heliconia buccinata Roxb. Hort. Beng. (1814) 19 (type!). 
Heliconiopsis amboinensis Miq. Fl. Ind. Bat. 3 (1858) 590. 
Folium mensarium album Rumph. Herb. Amb. 5: 142, t. 62, f. 2 
(sphalm. 1 in explic. p. 143). 

This species, a native of tropical America, must have been 



* Retained name, Vienna Code; Bihai Adans. (1763) is older. 



Z I N GIBERACE AE 



151 



introduced into Amboina at a comparatively early date, probably 
by the Portuguese colonists. In literature Folium buccinatum, 
figured on the same plate, is confused with Folium mensarium. 
The reduction to Musa bihai Linn, was made by Burman f ., Fl. 
Ind. (1768) 218, and to Heliconia bihai Linn, by Linnaeus, Mant. 
2 (1771) 211. The confusion between Folium mensarium and 
Folium buccinatum was apparently occasioned by Burman in 
writing the explanation of the plate, who reversed the names 
and figures. The latter is Cominsia gigantea (Scheff.) K. Sch. 
(see p. 167). 

The status of all the forms described in this chapter is now 
rather clear. Folium buccinatum asperum Rumph., Herb. Amb. 
5: t. 62, /. 1, is Cominsia gigantea (Scheff.) K. Sch.; the figure 
has been confused with fig. 2, Heliconia bihai Linn. Folium 
mensarium nigrum is, apparently, merely a variant of the preceding. 
Folium mensarium rubrum is a species of Cominsia, undescribed 
under the binomial system (see p. 168). Folium buccinatum 
album is Phacelophrynium robinsonii Val. 

ZINGIBERACEAE 
(By Th. Valeton) 
ZINGIBER Adanson 
ZINGIBER OFFICINALE Rose, in Trans. Linn. Soc. 7 (1807) 348. 
Amomum zingiber Linn. Sp. PI. (1753) 1. 

Zingiber majus album Rumph. Herb. Amb. 5: 156, t. 66, f. 1. 

Linnaeus made the first reduction of the Rumphian species 
to Amomum zingiber Linn., in Stickman Herb. Amb. (1754) 
20, Amoen. Acad. 4 (1759) 129, Sp. PI. ed. 2 (1762) 1, which, 
as Zingiber officinale Rose, is manifestly the correct disposition 
of it. The form described by Rumphius as Zingiber majus rubrum 
is probably merely a variant of the common ginger, Zingiber 
officinale Rose. 

ZINGIBER OFFICINALE Rose. var. MINOR Val. var. nov. 

Zingiber minus sive gramineum Rumph. Herb. Amb. 5: 161, t. 66, 
f. 2. 

Dimensione minore et proprietatibus rhizomatis insignis. 
Floribus paullum majoribus, labello orbiculari, lobis lateralibus 
late rotundato-ovatis basi subcordatis a genuina distinguenda. 

This variety is very commonly cultivated in western Java 
and occasionally produces flowers, although botanically it is 
quite unknown. It is easily identified by the native name sunti, 
mentioned by Rumphius and still in use, as well as by the 



152 



RUMPHIUS'S HERBARIUM AMBOINENSE 



description. Blume was wide of the mark in reducing it to 
Zingiber gramineum Blume, Enum. PL Jav. (1827) 45. The 
latter is a rather common wild species of Java and Cochin-China 
and is never cultivated. 

ZINGIBER ZERUMBET (Linn.) Smith Exot. Bot. 2 (1804) 103, t. 112. 
Amomum zerumbet Linn. Sp. PL (1753) 1. 

Lampujum majus domesticum Rumph. Herb. Amb. 5: 148, t. 6U, /. 1. 
Amboina, Toelehoe, Boerlage 2J*7 in herb, bog., July 16, 1900. 

This reduction of Lampujum majus domesticum was first 
made by Linnaeus, in Stickman Herb. Amb. (1754) 20, Amoen. 
Acad. 4 (1759) 129, Syst. ed. 10 (1759) 842, Sp. PL ed. 2 (1762) 
1, and is manifestly the correct disposition of it. Rumphius dis- 
tinguished a cultivated and a wild form, but he found no dif- 
ference between them, only that the latter flowered more 
abundantly. Blume did not think this was the genuine Z. 
zerumbet and reduced it to Z. amaricans BL; but his notions 
about true Z. zerumbet are not very clear, and specimens from 
Amboina {Boerlage 2J+7) in the Buitenzorg herbarium seem 
to agree with the type. Zingiber marginatum (Roxb.?) BL, 
which is quoted here by some authors, was based partly on 
Lampujum silvestre minus Rumph., Herb. Amb. 5: t. 61*, which 
is Globba marantina L., partly on an abnormal form of Z. 
zerumbet, with a central inflorescence instead of a lateral one, 
occurring sometimes in Java.* Zingiber montanum Link was 
based on Amomum montanum Koenig, in Retz. Obs. 3 (1779) 
51, quoted here by Koenig, 1. c, but Ridley, Fl. Mai. Penins. 2 
(1907) 28, thinks it represents Z. cassumunar Roxb. 

ZINGIBER ZERUMBET (L.) Smith, var. AMARICANS Val. 
Zingiber amaricans BL Enum. PI. Jav. (1827) 43. 
Lampujum minus Rumph. Herb. Amb. 5: 148. 

This reduction is based on Rumphius's assertion that this 
form was a native of Java, and thence brought to Amboina 
with the Malayan name lampujum ketjil or wangi. The latter 
is still very commonly used in Java for a form that is botanically 
not to be distinguished from lampujum pait y described by Blume 
as Zingiber amaricans. Blume's short description agrees wholly 
with the specimens I have examined. 

ZINGIBER CASSUMUNAR Roxb. in As. Research. 11 (1810) 347, t. 5. 
Bang leu m Rumph. Herb. Amb. 5 : 154, t. 65, f. 2. 

This reduction of Bangleum was apparently first made by 



* See van Zijp, Recueil Trav. Bot. Neerlandais 12 (1915) 346, t. S. 



ZING1BERACEAE 



153 



Blume, Enum. PL Jav. (1827) 42, which is manifestly the 
correct disposition of it. The figure represents a sterile speci- 
men only. 

ALPINIA Linnaeus 

ALPINIA GALANGA (Linn.) Sw. Obs. Bot. (1791) 8. 

Maranta galanga Linn. Sp. PI. ed. 2 (1763) 3 (type!). 
Galanga major Rumph. Herb. Amb. 5: 143, t. 63. 

This species is not represented in Robinson's collection. Ga- 
langa major Rumph. is the whole basis of Maranta galanga 
Linn., which typifies Alpinia galanga Sw. ; the plate is very 
generally cited in botanical literature under Alpinia galanga 
(Linn.) Sw. All the specimens I have seen from Amboina and 
Celebes (Boerlage, Teysmann, Riedel, Koorders) differ from 
the cultivated form in having pubescent leaves and inflores- 
cences. They probably represent the same form as the wild 
Javan one, Alpinia pyramidata Blume, considered by K. Schu- 
mann as a variety of Alpinia galanga Sw. Galanga minor 
Rumph., Herb. Amb. 5 : 144, t. 63, d, may represent a different 
species. 

ALPINIA GIGANTEA Blume Enum. PI. Jav. (1827) 59, non aliorum. 

Alpinia eremochlamys K. Sch. ex parte (specim. auth. in Herb. Bog.) 

in Engl. Pflanzenreich 20 (1904) 362, excl. descr. et fig. J+0, 0, citatis. 
Globba silvestris minor Rumph. Herb. Amb. 6: 141, excl. t. 63 citata 
(t. 62 et 63 ad Alpinia nutans spectant). 
Amboina, Hoetoemoeri road, Robinson PI. Rumph. Amb. 1J+1 (staminate), 
on a fern-covered hillside, altitude about 250 meters, the whole plant about 
6 meters high, locally known as geloba gar damn; Soja, Robinson PL 
Rumph. Amb. US (pistillate) on grassy slopes, altitude 375 meters, plant 
about 8 meters high. Collected also in Amboina by Teysmann at Hoetoe- 
moeri, and by Boerlage, No. 4-07, at Latoea, the latter recording the local 
names gardamoe oetan and anipa waccang. The Celebes form mentioned 
by Rumphius is probably a related but distinct species. 

K. Schumann, in Engl. Pflanzenreich 20 (1904) 355, regards 
the name Alpinia gigantea Blume as invalid on the basis : "nomen 
speciei mixtae melius deletur quam ulterius conservatur." Now 
Blume was wrong in referring Globba silvestris major to his 
species, but Alpinia gigantea Blume was based on a rather poor 
specimen and Reinwardt's manuscript, and hence it is not to be 
interpreted by the Rumphian synonym quoted by him. Robin- 
son referred his No. 1U1 to the form figured by Rumphius, 
t. 62, as Blume did, yet considering it as Globba silvestris minor. 
It is improbable that Rumphius would have depicted a fruiting 
specimen when he professes that this kind never bears fruit, 
while the figures distinctly show the withered bracts, which 



154 



RUMPHIUS'S HERBARIUM AMBOINENSE 



occur in Alpinia nutans and are very distinct in t. 63, 
fig. B, which I think is merely an almost life-sized branchlet of 
the form figured on the plate. According to Rumphius the 
racemes of Globba silvestris minor differ from Globba silvestris 
major in being longer, sterile, and without bracts, in all of 
which characters Robinson 1U1 agrees; the specimen presents 
only staminate flowers, with no vestiges of a gynaecium and no 
bracts. The specimen also agrees with Alpinia gigantea Blume 
in having very large pubescent leaves, so that Blume's species 
was not a "species mixta" at all and should be reinstated. The 
pistillate racemes of this plant, unknown to Rumphius, were also 
collected by Robinson, No. 1J+3, who did not, however, recognize 
that this form belonged with the one represented by his 141. In 
the Buitenzorg herbarium there are numerous specimens collected 
by Boerlage and by Teysmann presenting both pistillate and 
staminate flowers. 

ALPINIA NUTANS (Linn.) Rose, in Smith Exot. Bot. 2 (1805) quoad 
syn. Linn., excl. t. 160 et descr. quae est A. speciosa K. Sch. 
Globba nutans Linn. Mant. 2 (1771) 170 (type!). 

Alpinia papuana K. Schum. in Engl. Pflanzenreich 20 (1904) 355, 

p. p., quoad specim. Moluccana. 
Alpinia moluccana Gagnep. in Bull. Soc. Bot. France 48 (1902) 90. 
Alpinia eubractea Val. (non K. Schum.) Ic. Bogor. 3 (1909) t. 300. 
Alpinia gigantea Val. (non Blume) Nova Guinea 8 (1913) 943. 
Globba silvestris major Rumph. Herb. Amb. 6: 140, t. 62, 63 (in 

explicatione tab., t. 63, sphalmate pro Globba silvestris minor dis- 

tincta habetur). 

Amboina, Lateri, Robinson PL Rumph. Amb. 142, September 5, 1913, in 
forests, altitude about 275 meters, locally known as geloba merah. Also 
collected by Teysmann in the same locality, and by Boerlage 83 p. p. 

This species is widely distributed in the Eastern Archipelago 
and has been described under various names, but Globba nutans 
Linn., the name-bringing synonym of Alpinia nutans Rose, was 
based wholly on Globba silvestris major Rumph., Alpinia nutans 
Rose, thus being the oldest name for the species. The younger 
Linnaeus, Suppl. (1781) 79, followed by Giseke, Prael. Ord. 
Nat. PL (1792) 251, erroneously referred both of the Rumphian 
illustrations to Renealmia exaltata Linn, f ., but the plant actually 
described under this name was from South America and is 
entirely different from the Moluccan one. By some authors, 
Murray, Syst. (1774) 67, and Richter, Codex Bot. Linn. (1840) 
42, the figures are erroneously cited as t. 12 and t. 13, instead of 
t. 62 and t. 63. Alpinia nutans Rose, as actually described 
and figured, is Alpinia speciosa (Wendl.) K. Schum. Alpinia 
nutans K. Schum., non Rose, is Alpinia oceanica Burkill. AU 



ZINGIBERACEAE 



155 



pinia moluccana Gagnep. is exactly identical with Alpinia nutans 
as here interpreted. Alpinia papuana Scheff., in Ann. Jard. 
Bot. Buitenz. 1 (1876) 56 (type in Herb. Hort. Bog.), has a 
different calyx and much smaller flowers and inflorescences than 
the present species. Alpinia eubractea K. Sch. (non Val.) from 
Celebes (Sarasin 8J>6) , differs conspicuously in its long, narrow, 
truncate bracteoles.* Alpinia colossea K. Schum. is the same as 
A. papuana Scheff. Alpinia gigantea Blume is definitely refer- 
red by K. Schumann to Scheffer's species, although at the same 
time he rebukes Blume for quoting Rumphius's t. 62 instead of 
t. 63 and for describing the leaves as pubescent instead of 
glabrous. Following K. Schumann's interpretation of Blume's 
species, I adopted the latter's specific name for the wrong species 
(Nova Guinea 1. c.) . If Blume had meant this species, he surely 
would have quoted t. 63 and would also have mentioned the 
striking bracts (which are shown, but much less conspicuously, 
in t. 62), and he would have described the leaves as glabrous. 
Blume's description certainly applies to a different species; see 
above. As already noted, Globba nutans Linn, supplies the 
oldest specific name for this species, but which was rejected 
by K. Schumann because Linnaeus quoted both t. 62 and t. 63, 
which he, K. Schumann, considered to represent distinct species ; 
yet in this K. Schumann was wrong and Linnaeus was right, 
as both figures manifestly apply to but a single species. 

ALPINIA MALACCENSIS Rose, in Trans. Linn. Soc. 8 (1807) 330. 

Galanga malaccensis Rumph. Herb. Amb. 5: 176, t. 71, f. 1 (incl. 
Bangle malacca) . 

The Rumphian reference cited above was placed by Burman f ., 
PI. Ind. (1768) 2, under Maranta malaccensis Burm. f., which is 
the name-bringing synonym of Alpinia malaccensis Rose. 
Roxburgh, Asiat. Res. 11 (1810) 353, referred to Alpinia 
malaccensis material from Chittagong, and his interpretation 
has been accepted by most authors. Blume, Enum. (1827) 59, 
without mention of Roxburgh's description, interpreted the 
common Javan form as Burman's species, and his reduction and 
diagnosis have been overlooked by all authors, including K. 
Schumann. Ridley, who collected the Javan form in Malacca 
denied that it is identical with Roxburgh's interpretation of 
Alpinia malaccensis and redescribed it as Alpinia nobilis Ridl., 
in Journ. Roy. As. Soc. Straits Branch 32 (1899) 169, without 
mention of Alpinia malaccensis Rose. It seems to me more 
probable that Blume's interpretation of the Rumphian species is 



* See K. Schumann in Engl. Pflanzenreich 20 (1904) t. M, f. A. 



156 



RUMPHIUS'S HERBARIUM AMBOINENSE 



the correct one as Rumphius states that the species occurred 
most commonly in Java and in Malacca ; it is more probable that 
an Amboina species would also occur in Java than in Chittagong. 
The Rumphian description agrees equally well with both forms, 
and as this particular one has not been collected in Amboina 
since Rumphius's time, I cannot determine, with certainty, 
whether or not Alpinia malaccensis Rose, is the same as A. 
nobilis Ridl. Roscoe's species will have to remain among those 
of uncertain status until Amboina specimens are available for 
study. 

ALPINIA UVIFORMIS (Linn.) Horan. Monogr. (1862) 35 (type!). 
Globba uviformis Linn. Mant. 2 (1771) 171 (type!). 
Giobba uviformis Rumph. Herb. Amb. 6: 138, t. 59, f. 2. 

A species of doubtful status, known only from Rumphius's 
figure and description. K. Schumann places it in the section 
Cylindrostachys. However, it has every appearance of a 
Plagiostachys, and I suspect that it really belongs in this genus, 
not in Alpinia. The figure and the description are rather good. 
It differs from all known species of Plagiostachys, this genus 
extending from Malacca to Banka, Borneo, and the Philippines, 
in its much longer, white flowers and tomentose fruits. It is 
singular that this species, reported by Rumphius from Amboina 
and Celebes and stated by him to be so common in southern 
Celebes that the Buginese use it for thatching, has not been 
collected by any modern botanist or collector. 

ALPINIA sp. 

Globba silvestris pada kanka Rumph. Herb. Amb. 6: 142. 

This was described from specimens from eastern Celebes, 
there known as padakanka. The characters indicated — terminal 
raceme, red flowers, and small globose fruits — point to Alpinia. 
Hasskarl thought that it might be near Alpinia pyramidata 
Blume. 

RIEDELIA Oliver 

RIEDELIA LAN AT A (Scheff.) K. Schum. msc. in Herb. Bog.; Valeton in 
Ic. Bogor. 4 (1913) t. 373, Nova Guinea 8 : 961. 
Hedychium lanatum Scheff. in Ann. Jard. Bot. Buitenz. 1 (1876) 57. 
Riedelia curviflora Oliv. in Hook. Ic. PI. 15 (1883) t. 1419; K. Sch. 

in Engl. Pflanzenreich 20 (1904) 374 p. p. 
Thylacophora pogonocheilus Ridl. in Trans. Linn. Soc. Bot. 9 (1916) 
210, t. 5, f. 74, t. 6, f. 75-82. 
■ ? Globba lawassi Malacca Rumph. Herb. Amb. 6: 139. 
Amboina, Lateri, Robinson PI. Rumph. Amb. 54.3, in light forests, 
altitude about 35 meters, locally known as globba baubau. 

This species has not been previously collected in Amboina, 



ZINGIBERACEAE 



157 



although it is well known in Boeroe (Riedel, Teysmann) and 
occurs in New Guinea (Teysmann, Versteeg). I reduce to it 
Globba lawassi Malacca of Rumphius, because the fruits agree 
with Rumphius's description, while the seeds have a marked 
flavor of cardamon. 

A MOM U M sensu Benthamiano * 

AMOMUM CARDAMOMUM Willd. (not Linn, which is Elettaria cardamo- 
mum) Sp. PL 1 (1797) 8; Roxb. PL Coromandel 3: 22, t. 227, PI. 
Ind. ed. 2, 1 (1832) 37; K. Schum. in Engl. Pflanzenreich 20 (1904) 
238 (Amomum cardamon) . 
Cardamomum minus Rumph. Herb. Amb. 5 : 152, t. 65, f. 1. 

This was reduced by Linnaeus to Amomum cardamomum 
Linn., in Stickman Herb. Amb. (1754) 20, Amoen. Acad. 4 
(1759) 129, but is entirely different from the Linnean species 
as originally published in 1753, which is Elettaria cardamomum 
Maton. It is not represented by botanical specimens from Am- 
boina, but occurs in Java, cultivated and spontaneous, and in 
Sumatra. 

In the same chapter Rumphius briefly mentions two other 
species, the fruits of which were considered as kinds of car- 
damon. He distinguishes in all three kinds of cardamon, none 
of which is a native of Amboina, as follows : 

Cardamom majus "the round Cardamom of Java." 

This is, according to Pereira, who compared fruits known by this name 
with authentic specimens in Kew, doubtless Amomum maximum Roxb. PL 
Ind. 1 (1820) 41, a native of Java [Pereira Materia medica 2 (1849) 106]. 
Cardamom medium sive minus —Amomum cardamomum Willd. 

This is the only one that was cultivated, though with little success, 
in Amboina. 

Cardamom verum fructibus minimis tr\gor\\s = Elettaria cardamomum 
Maton. 

AMOMUM ROSEUM (Teysm. & Binn.) Benth. & Hook. f. ex Jackson Index 

Kewensis (1895) 108, non Roxb. 
Donacodes rosea Teysm. & Binn. Cat. Hort. Bogor. (1866) 58. 
Amomum truncatum Gagnep. in Bull. Soc. Bot. Fr. 51 (1903) 164. 
Globba crispa II rubra Rumph. Herb. Amb. 6: 137, t. 60, /. B, C, D, 

t. 61, f. 2. 

Amboina, Koesoekoesoe sereh, Robinson PL Rumph. Amb. 140, August 
23, 1913, locally known as gelobba; Teysmann. Celebes, Koorders 592c, 
19651$. 

Loureiro, Fl. Cochinch. (1790) 4, referred this to Amomum 
villosum Lour., that is, Globba crispa Rumph. t. 61, but mani- 

* This genus in its widest sense contains species placed by different 
authors under Elettaria, Geanthus, Donacodes, Phaeomeria, Hornstedtia, 
and Amomum. 



158 



RUMPHIUS'S HERBARIUM AMBOINENSE 



festly the Amboina plant is entirely different from the Cochin- 
China one that Loureiro described. In this reduction Loureiro 
was followed by numerous authors. Hasskarl, Neue Schliissel 
(1866) 174, thought that it might be Amomum aculeatum Roxb., 
which can scarcely be the correct disposition of it. Donacodes 
rosea Teysm. & Binn. was based on Amboina specimens, and 
our material agrees closely with the figure given by K. Schumann 
in Engler's Pflanzenreich 20 (1904) 230. The Celebes speci- 
mens described by Gagnepain as Amomum truncatum Gagnep. 
are certainly conspecific with the Amboina ones, although they 
have glabrous leaves, while the Amboina specimens have long- 
ciliate leaves. 

AMOMUM ACRE Val. sp. nov. 

Globba acris Rumph. Herb. Amb. 6: 140. 

Folia? Flores? Ovarium ellipticum sericeum sublaeve. In- 
florescentia fructifera racemosa simplex elongata densiflora. 
Rachis valida 10 ad 12 mm crassa. Bracteae persistentes co- 
riaceae oblongae versus basin attenuatae, inferiores 30 ad 35 
mm longae 8 ad 10 mm latae valde scariosae versus apicem 
racemi sensim minores. Bracteolae binae approximatae prope 
basin pedicelli intervallo brevi insertae ovatae acutae concavae 
dorso leviter carinatae, valde coriaceae sericeo-tomentosae, 12 
ad 15 mm longae 6 ad 8 latae. Pedicelli validi 15 ad 20 mm 
longi, 2 ad 3 mm crassi bracteis breviores, sericei. l^ructus late 
ellipsoidei (30X20 mm) utrinque attenuati, apice breviter 
rostrati et in calycem coriaceum persistentem ad basin usque 
trifidum, segmentis 10 ad 15 mm longis apice spinulosis, ex- 
currentes, pericarpio carnoso-coriaceo spinis compressis e basi 
lata acute trigonis 1 ad 2 mm altis, verticaliter subseriatis haud 
densis obsesso, triloculares septis tenuibus seminibus numerosis 
arillatis conglutinatis. 

This species is very closely allied to Amomum roseum (T. & 
B.) Benth. & Hook. f. (see p. 157). The infructescence of the 
latter resembles very much that of Amomum acre, but is smaller 
in all parts except the bracts, which are longer, not scariose; 
and end in a spinous point. Also the spines of the fruit are 
much denser near the apex of the fruit. This species is repre- 
sented in the Buitenzorg herbarium by fruiting specimens from 
Macassar, Celebes, where the plant still bears the Malay name 
panas, as cited by Rumphius. It seems to be near Amomum 
roseum Benth. & Hook. f. Hasskarl, Neue Schliissel (1866) 
175, thought that it might be Amomum uliginosum Koenig; 
this is manifestly an erroneous reduction of it. 



ZINGIBERACEAE 



159 



? AM O MUM ACULEATUM Roxb. Asiat. Research. 11 (1810) 344, t. 6; 
Blume Enura. PL Jav. (1827) 50; K. Sch. in Engl. Pflanzenreich 
20 (1904) 240. 

Amomum hatuanum Naves Novis. App. Fl. Filip. (1880) 224 (type!). 
? Globba hatuana Rumph. Herb. Amb. 6: 138. 

Amomum aculeatum Roxb. is common in Java, where it grows 
both wild and cultivated, and I think that it occurs in various 
forms or Varieties throughout the Eastern Archipelago. 
Among the Papuan forms, which I have described as varieties of 
this species, Amomum gymnocarpum Val., Nova Guinea 8: 926, 
apparently agrees with Rumphius's description. Rumphius's 
description is the whole basis of Amomum hatuanum Naves, 
a name not listed in Index Kewensis. 

Globba longa minor Rumph., Herb. Amb. 6: 136, a native of 
Java, is probably Hornstedtia minor (Blume) K. Sch. (Elettaria 
minor Blume). 

? AMOMUM MAGNIFICUM (Rose.) Benth. & Hook. f. Gen. PI. 3 (1883) 
644. 

Alpinia magnified Rose. Monandr. PI. (1828) t. 75. 
Phaeomeria magnified Lindl. Introd. Nat. Syst. ed. 2 (1835) 466. 
Nicolaia imperialis Horan. Monogr. (1862) 32, t. 1. 
Hornstedtia imperialis Ridl. in Journ. Roy. As. Soc. Str. Branch 32 
(1899) 148. 

Nicolaia magnified K. Schum. in Engl. Bot. Jahrb. 27 (1899) 303; 

Val. in Bull. Inst. Bot. Buitenz. 20 (1904) 35. 
Globba silvestris sulica Rumph. Herb. Amb. 6: 141. 

Rumphius's description applies unmistakably to Nicolaia, and 
Hasskarl, Neue Schlussel (1866) 175, considered it to be Nico- 
laia speciosa Horan. or an allied species. I have placed it 
under Amomum magnificum with doubt, because I suppose the 
Celebes specimens collected by Sarasin, cited by K. Schumann 
under Nicolaia magnified K. Schum., to represent Rumphius's 
species, which was from Celebes. Whether the form Rumphius 
described is Nicolaia magnifica K. Schum., N. hemisphaerica 
Horan., or N. speciosa Horan., I do not venture definitely to 
decide. I consider N. speciosa Horan. to be distinct from N. 
magnifica K. Schum. on account of the color of the labellum. 

AMOMUM RUMPHII Smith in Rees Cyclop. 39 (1819) no. 15 (type!). 
Elettaria musacea Horan. Monogr. (1862) 31 (type!). 
Donacodes incarnata T. & B. Cat. Hort. Bogor. (1866) 380. 
Hornstedtia elongata K. Schum. in Engl. Pflanzenreich 20 (1904) 
192 p. p. excl. descriptione, quae refert partim ad Amomum max- 
imum Bl. !, partim ad Hornstedtiae spec, incertam. 
Globba longa s. vulgaris Rumph. Herb. Amb. 6: 134, t. 60. 
Amboina, Hatiwe and Batoe gadjah, Robinson PI. Rumph. Amb. HU, 



160 



RUMPHIUS'S HERBARIUM AMBOINENSE 



August and September, 1913, on wooded hillsides and river banks, altitude 

20 to 250 meters; Teysw.ann. Cultivated in the botanic garden at Bui- 
tenzorg, from Amboina, under the native name geloba. 

A critical examination of two authentic specimens in the 
Buitenzorg herbarium labelled by K. Schumann "Bearbeitet fur 
das Pflanzenreich, ,, both from plants cultivated in the botanic 
garden, shows them to be different from each other. One of 
them is wrongly labelled "Donacodes minor" and is apparently 
D. elongata T. & B. The other is the type specimen of D. 
incarnata T. & B., a native of Amboina; this is identical with 
Robinson's Amboina material and is easily distinguished from 
D. elongata T. & B. by its densely tomentose ligule. Donacodes 
elongata T. & B. and D. incarnata T. & B. are nomina nuda, 
and the latter is apparently identical with Globba longa Rumph. 
Older names are Amomum rumphii Smith and Elettaria musa- 
cea Horan., both based on Rumphius. K. Schumann was wrong 
in referring Globba longa Rumph. to Amomum TQseum (T. & 
B.) Benth. & Hook. f. 

AMOMUM sp. 

Globba crispa I viridis Rumph. Herb. Amb. 6: 137, t. 61, f. 1. 

This Rumphian species has not been located, yet the descrip- 
tion and figure are so good that it could not readily be overlooked 
had a plant presenting its characters appeared in modern collec- 
tions from the Moluccas. Willdenow, Sp. PL 1 (1797) 8, after 
Koenig, placed the Rumphian species as a synonym of Amomum 
echinatum Willd. Willdenow's species was based on Koenig's 
Amomum 2 or Globba crispa, a native of Malacca; the latter 
has tubercied, not aculeate, fruits and a filiform rhizome and 
differs totally from Globba crispa Rumph. Amomum echinatum 
Baker, of Ceylon, described also by Trimen under this name, is 
still another quite different species, with copious curved spines. 
Loureiro, Fl. Cochinch. (1790) 4, referred Globba crispa Rumph. 
to Amomum villosum Lour., where it manifestly does not belong. 

AMOMUM sp. 

Globba silvestris subterranea Rumph. Herb. Amb. 6: 142. 

Amomum hochreutineri VaL, a species with white subter- 
ranean fruits, as described by Rumphius, occurs in the mountains 
of western Java. It is allied to Amomum hypoleucum Thw. and 
like the latter is characterized by the silvery lower surface of the 
leaves ; it is assumed that Rumphius would have mentioned this 
character had he intended this species. Hasskarl thought that 
it might be near Amomum gracile Blume. 



ZINGIBERACEAE 



161 



AMOMUM sp.? 

Globba sekala Rumph. Herb. Amb. 6: 141. 

This herb, a native of the eastern coast of Celebes, is said 
to resemble Alpinia gigantea in habit, but to be smaller and 
with reddish flowers issuing from the rhizome as in Amomum 
roseum. The species has not been found again so far as can 
be ascertained. It might be Alpinia melichroa K. Schum., in 
Engl. Pflanzenr. 20 (1904) 364, collected by Sarasin (No. 871) 
in Celebes. Nothing definite can be stated regarding its proper 
position. 

HEDYCHIUM Koenig 

HEDYCHIUM CHRYSOLEUCUM Hook. f. in Curtis's Bot. Mag. 76 (1850) 
t. 4516; K. Schum. in Engl. Pflanzenreich 20 (1904) 46. 
Gandasulium Rumph. Herb. Amb. 5: 175, t. 69, f. 3. 
Amboina, Robinson PI. Rumph. Amb. 1^5, August, September, and 
October, 1913, from cultivated plants, locally known as gandasuli. 

Gandasulium Rumph. was reduced by Koenig to Hedychium 
coronarium Koenig in the original description of that species, 
and this disposition of it has been accepted by all subsequent 
authors. It is not certain, however, that Gandasulium Rumph. 
is identical with Hedychium coronarium Koenig, which was 
based on Malacca specimens. Koenig's original description is 
not sufficiently detailed to show its exact status. If the Am- 
boina material cited above really represents Gandasulium 
Rumph., as I suppose it does, it is certainly not the same as 
Hedychium coronarium as considered by Roscoe, Monandr. PL 
t. 51, and by K. Schumann, in Engl. Pflanzenreich 20 (1904) 44. 
It seems to me to agree perfectly with the figure of Hedychium 
chrysoleucum Hook. f. quoted above, the drawing having been 
made from specimens sent by Roxburgh; it is not impossible 
that Roxburgh's plants came originally from Amboina, although 
it seems not to be a native of Amboina. It occurs in Java only 
as a cultivated plant. 

KAEM PFERI A Linnaeus 

KAEM PFERI A GALANGA Linn. Sp. PI. (1753) 2. 

Soncorus Rumph. Herb. Amb. 5: 173, t. 69, f. 2. 

Not represented in our Amboina collections. Linnaeus orig- 
inally reduced Soncorus Rumph. to Kaempferia galanga Linn., 
in Stickman Herb. Amb. (1754) 20, Amoen. Acad. 4 (1759) 
129, Syst. ed. 10 (1759) 843, Sp. PL ed. 2 (1762) 3, which is 

144971 11 



162 



RUMPHIUS'S HERBARIUM AMBOINENSE 



apparently the correct disposition of it and one that has been 
accepted by all authors. 

KAEM PFERI A PAN DU RATA Roxb. in As. Research. 11 (1810) 328, t. 2. 
Gastrochilus panduratum Ridl. in Journ. As. Soc. Str. Branch 32 
(1899) 114. 

Zerumbed claviculatum Rumph. Herb. Amb. 5: 172, t. 69, f. 1. 

Linnaeus originally reduced Zerumbed claviculatum Rumph. 
to Kaempferia rotunda Linn., in Stickman Herb. Amb. (1754) 
20, Amoen. Acad. 4 (1759) 129, which is manifestly an er- 
roneous disposition of it. Roxburgh, Fl. Ind. 1 (1820) 18, 
placed it under Kaempferia pandurata Roxb., which was de- 
scribed from Sumatran specimens. This reduction has been 
accepted by Roemer and Schultes, Blume, Dietrich, Miquel, and 
Horaninow and is apparently the correct disposition of it. 

GLOBBA Linnaeus 

GLOBBA MARANTINA Linn. Mant. 2 (1771) 170. 

Lampujum silvestre minus Rumph. Herb. Amb. 5: 150, t. 6U, f. 2. 
Amboina, Soja, Robinson PL Rumph. Amb. 145, August, 1913, along 
roadsides and on banks, altitude 5 to 400 meters, locally known as kuning 
utan. 

Rumphius's figure is a very poor one, but it is certainly refer- 
able to Globba marantina Linn. This reduction seems first to 
have been made by Roemer and Schultes, Syst. Mant. 1 (1822) 
46. 

CURCUMA Willdenow 

Rumphius described, in all, nine forms of Curcuma, of which 
he declares that the first three, considered in chapter sixteen, 
belong to the true Curcuma. In the following chapter he deals 
with a group of species, which he considered under the "generic" 
name Zerumbed, on account of the assertions of the natives, who 
deliberately distinguish these forms from the true Curcuma, or 
koenjit, uniting them under a proper group name, tommon. 
Rumphius confessed, however, that he himself was not able to 
find characters by which the two groups could be distinguished. 
However, if one compares the native names cited by Rumphius 
for the six species mentioned under Zerumbed, with those now 
used in Java, one is impressed by the fact that, with a single 
exception, all of them are applied to species belonging in the 
section Exantha; while the three species placed by Rumphius 
under the name Curcuma, all belong in the section Mesantha. 
The clue to Rumphius's enigma apparently lies in the fact that 



ZINGIBERACEAE 



163 



he had not seen the flowers of most of the species, and hence 
did not know of the peculiar distinguishing character of the 
inflorescence. This may also explain why t. 68 was placed under 
Zerumbed, rather than under Curcuma, where it really belongs. 

Of the six species enumerated in chapter seventeen, under 
Zerumbed, the first and second are discussed below; the third 
(album), fourth (giring), and sixth (manga), all cultivated in Java, 
are still undescribed under the binomial system; while the fifth 
(frigidum) , as regards the name ties, but not the description, 
belongs to Curcuma viridiflora Roxb. Fl. Ind. 1 (1820) 34, a 
species of the section Mesantha. Roxburgh himself reduced 
the fourth species to Curcuma viridiflora Roxb., but that was 
evidently an error on his part. 

CURCUMA LONG A (non Linn, quae est prob. C. aromatica Salisb., vide 
Hermann Hort. Acat. Lugd. Bot. Descr. (1687) 208, t. 209) Koenig 
in Retz. Obs. 3 (1783) 72, non alio rum. 
Curcuma domestica minor (et major?) Rumph. Herb. Amb. 5: 164, 
t. 67. 

Rumphius's description agrees exactly with our commonly 
cultivated curry plant, as do the Malay names cited by him, 
and the form that he described and figured is apparently 
identical with that described by Koenig as Curcuma longa. It 
is characterized by its central spike, white bracts, and orange 
rhizomes. In Java a tall form and a short form occur, but I 
cannot detect any notable differences between them. Curcuma 
longa Linn., Sp. PI. (1753) 2, was based wholly on Curcuma 
radice longa Herm., which probably is the same as Curcuma 
aromatica Salisb. Linnaeus, however, reduced the Rumphian 
species, as figured, to Curcuma longa Linn., in Stickman Herb. 
Amb. (1754) 20, Amoen. Acad. 4 (1759) 129, Syst. ed. 10 (1759) 
843; while in the second edition of the Species Plantarum (1762) 
3, he erroneously referred it to Curcuma rotunda Linn., which is 
supposed to be the same as Kaempferia pandurata Roxb. 

CURCUMA PETIOLATA Roxb. Fl. Ind. 1 (1820) 37, sensu latissimo K. 
Schum. in Engl. Pflanzenreich 20 (1904) 102. 
Curcuma agrestis Rumph. Herb. Amb. 5: 164. 

This species, in the wide sense that it was interpreted by K. 
Schumann, was until recently known only from Pegu and 
Moulmein, but has been detected growing also in Java. Rumph- 
ius's description agrees so closely with that of Roxburgh, 
that I do not hesitate in reducing Curcuma agrestis Rumph. to 
C. petiolata Roxb. 



164 



RUMPHIUS'S HERBARIUM AMBOINENSE 



CURCUMA ZEDOARIA (Berg.) Rose. Monandr. PI. (1828) t. 109. 
Amomum zedoaria Berg. Mat. Med. (1788) 41. 

Zerumbed majus Rumph. Herb. Amb. 5: 168, excl. t. 68 citata (quae 
prob. ad C. viridifloram Roxb. pertinet). 

Zerumbed of Rumphius was first placed by Linnaeus under 
Curcuma rotunda Linn., in Stickman Herb. Amb. (1754) 20, 
Amoen. Acad. 4 (1759) 129, Syst. ed. 10 (1759) 842, but is 
manifestly Curcuma zedoaria Rose, where it was placed by 
Willdenow, with doubt (sub Amomum zedoaria Berg.), and 
by Roemer and Schultes. It has, by various authors, been placed 
under Costus arabicus Linn., Amomum latifolium Lam., Amo- 
mum zerumbeth Koenig, Renealmia exaltata Linn., and Curcuma 
zerumbet Roxb. 

CURCUMA AERUGINOSA Roxb. Fl. Ind. 1 (1820) 27; Rose. Monandr. 

PI. (1828) t. 106. 
Zerumbed nigrum Rumph. Herb. Amb. 5: 169. 

This form occurs in Java, cultivated and wild, and, according 
to Roxburgh, in Burma. Roxburgh placed Zerumbed nigrum 
Rumph. under Curcuma caesia Roxb., but the Javan plant, which 
is certainly the same as the one that Rumphius described, agrees 
better with Curcuma aeruginosa Roxb. 

COSTUS Linnaeus 

COSTUS SPECIOSUS (non Smith) Blume Enum. (1827) 61, non Banksia 
speciosa Koenig. 

Costus speciosus var. glabra K. Sch. in Engl. Pflanzenreich 20 (1904) 
398. 

Herba spiralis II laevis Rumph. Herb. Amb. 6: 143, t. 6Jf, f. 2. 
Amboina, Soja, Robinson PI. Rumph. Amb. U7, August 14, 1913, in 
forests, altitude about 400 meters, flowers white, fruit red. 

COSTUS SPECIOSUS Blume var. HIRSUTUS Blume Enum. (1827) 61; 
Val. in Nova Guinea 8: 982. 
Costus speciosus var. lasiocalyx K. Sch. in Engl. Pflanzenreich 20 
(1904) 398. 

Costus speciosus var. argyrophyllus Ridl. in Herb. Kds., non Ridl. in 

Mat. Fl. Mai. Pen. 2 (1907) 24, which is C. sericea Blume. 
Herba spiralis I hirsuta Rumph. Herb. Amb. 6: 143, t. 16U, /• -Z. 

This has been collected in Amboina by Boerlage and is known 
from Java, Celebes, Key, and New Guinea. Koenig originally 
reduced Herba spiralis Rumph. to Costus speciosus Sm., in which 
he has been rather uncritically followed by all recent authors. 
The Malacca form, typical Costus speciosus Sm., which is com- 
monly cultivated in Batavia, but which is not wild in the Archi- 



CANNACEAE 



165 



pelago, has flowers four times as large as the Javan and Amboina 
forms, which again differ from each other. Besides Costus 
speciosus I think there are two other known species common in 
the Malayan region. One is Costus sericeus Blume (argyro- 
phyllus Ridl.), the other the form represented by Rumphius's 
figure, which will probably need a distinctive specific name. 

TAPEINOCHILUS Miquel 

TAPEINOCHILUS AN AN ASS A E (Hassk.) K. Schum. in Engl. Bot. Jahrb. 
27 (1899) 349; Engl. Pflanzenreich 20 (1904) 436. 

Costus ? ananassae Hassk. in Abhandl. Naturf. Gesellsch. Halle 9 

(1866) 335 (Neue Schliissel 191) (type!). 
Tapeinochilus pungens Miq. in Ann. Mus. Bot. Lugd. Bat. 4 (1868-69) 

101, t. 4. 

Tubu tubu Rumph. Herb. Amb. 7: 52, t. 22, f. 2. 

The original description of the species proposed by Hasskarl 
was based entirely on Rumphius. The species was not published 
by Hasskarl under Tapeinochilus as indicated by K. Schumann, 
but under Costus. Neither of the names appears in Index Kew- 
ensis of in its supplements to date. 

CANNACEAE 

CANNA Linnaeus 

CANNA INDICA Linn. Sp. PL (1753) 1. 

Cannacorus Rumph. Herb. Amb. 5: 177, t. 71, f. 2. 

Amboina, Roemah tiga and about the town of Amboina, Robinson PI. 
Rumph. Amb. 410, August 15, 1913, along river banks, in sago swamps, etc., 
locally known as tasibe mera. 

This is the common, spontaneous, Indo-Malayan form with 
small red flowers that is generally placed under Carina indica 
Linn. As noted by Kranzlin, in Engl. Pflanzenreich 56 (1912) 
59, it is very difficult to determine the exact status of Canna 
indica, as described by Linnaeus. The Linnean species, as 
here interpreted, is apparently not the form described by 
Kranzlin. Cannacorus was originally reduced by Linnaeus to 
Canna angustifolia Linn., in Stickman Herb. Amb. (1754) 20, 
Amoen. Acad. 4 (1759) 130, Syst. ed. 10 (1759) 842, which is 
certainly an error; in the second edition of his Species Planta- 
rum (1762) 1, however, he placed Cannacorus under Canna 
indica Linn., which seems to be the correct disposition of it. 
By other authors it has been placed under Canna patens Rose, 
C. orientalis Rose, and C. coccinea Ait. 



166 



RUMPHIUS'S HERBARIUM AMBOINENSE 



MARANTACEAE 
(By Th. Valeton) 
DON AX Loureiro 

DON AX CAN N I FORM IS (Forst.) K. Schum. in Engl. Bot. Jahrb. 15 (1893) 
440; Rolfe in Journ. Bot. 45 (1907) 243. 
Thalia cannaeformis Forst. Prodr. (1786) 1. 

Actoplanes canniformis K. Schum. in Engl. Pflanzenreich 11 (1902) 
34 p. p., excl. syn. Maranta grandis Miq. quae est A. ridleyi K. 
Schum. 

Arundastrum Rumph. Herb. Amb. 4: 22, t. 7. 
Amboina, Soja and Way tommo, Robinson PI. Rumph. Amb. 520, August, 
1913, along river banks and in forests, altitude 45 to 400 meters; between 
Wae and Toelehoe and at Laha, Boerlage 191, 516, in Herbi Bog. 

Arundastrum Rumph. was originally reduced to Maranta 
arundinacea Linn., as a variety, by Lamarck, Encycl. 2 (1788) 
588, where it manifestly does not belong. Loureiro, Fl. 
Cochinch. (1790) 11, placed it under Donax arundastrum Lour., 
which is a species of Donax distinct from D. canniformis K. 
Schum. By Willdenow, Sp. PL 1 (1797) 13, it was erroneously- 
placed under the American Maranta tonckat Aubl.= Stromanthe 
tonckat Eichl. Roxburgh, Fl. Ind. ed. 2, 1 (1832) 2, placed it 
under Phrynium dichotomum Roxb., which is Schumannianthus 
dichotomus Gagnep.; and Dietrich placed it under Maranta 
dichotoma Wall., a synonym of the latter species. The synonymy 
of Donax canniformis K. Schum. and of the several species in 
the allied genus Schumannianthus is very complicated, but has 
been admirably cleared up by Rolfe in Journ. Bot. 45 (1907) 
242-244. 

PH ACELOPH RYN I U M K. Schumann 
PH ACELOPH RYN I U M ROBINSONII Val. sp. nov. 

Folium buccinatum album Rumph. Herb. Amb. 5: 142. 
Amboina, Mahija, Robinson PL Rumph. Amb. 521 (type), August 7, 
1913, along small streams, altitude about 250 meters, locally known as 
dauri; Waigama, Teysmann sine numero in Herb. Bog. (specim. fructif.). 

Herba bimetralis. Caulis florifer 150 ad 450 mm longus. 
Folia longissime petiolata (pet. propr. 100 mm, vagina fere 
1,200 mm), maxima late oblonga vel elliptica (lamina 400x250, 
vel 470x240, vel 500x170 mm) apice attenuata vel subrotun- 
data breviter (10 mm) apiculata, basi vulgo rotundata ad pe- 
tiolum constricta acuta, glaberrima. Inflorescentiae pedunculo 
100 ad 200 mm longo, rhachi glabra, singulae foliis 1 ad 2 
comitatae subsimplices, haud interruptae, ramis paucis (2 ad 
4) e bracteis inferioribus exortis erectis, spicae primariae bre- 
viores, internodia inferiora 30 ad 65 mm longa, superiora brevia 



MARANTACEAE 



167 



(5 ad 16 mm), bracteis spicularum densis plus minus occulta. 
Bracteae primariae 45 ad 25 mm longae, oblongae mucronatae 
superiores minores acutae ad 20 mm longae apice et basi minute 
puberulae, erectae. Florum paria 2 vel 3 bracteolis teneris 
paucis. Fl. brevipedicellati ad 25 mm longi. Ovarium teres 
(3 mm) appresse villosum, triloculare, triovulatum. Sepala 
anguste lanceolata apice acuta incurva 7 mm longa. Floris 
tubus gracilis elongatus (16 mm longus), petala lineari-oblonga 
genitalia superantia circ. medio tubo inserta ±10 ad 12 mm 
longa. Staminodium exterius singulum valde elongatum lineare 
7 mm longum, 1.5 mm latum, labellum (std. callosum) elongatum 
ellipticum apice dilatatum emarginatum valde concavum, callo 
superne tenui glabro, lobulo basali velutino (±6 mm longum). 
Cucullum multo brevius, appendicula haud magna, basi auri- 
culata munitum, stamen fertile subfiliforme, loculo petaloideo 
angusto inconspicuo 4 mm longum. Caryopsis parva (10 mm 
longa, 5 mm lata) teres oblonga apice rotundata et calyce 
coronata, uni latere concava, pericarpio tenui stramineo appresse 
hirsuto, abortu nunc unilocularis indehiscens, monospermus 
(teste Rumphius dispermus). Semen teres apice acutum, basi 
arillo magno nigro bipartito lobis magnis attenuatis instructum. 
Embryo leviter curvatum. 

Doctor Robinson's hypothesis that the specimen of his collec- 
tion, cited above, represented Folium buccinatum album Rumph. 
was a very ingenious one. Rumphius's description tallies so 
exactly with the specimens that it would be impossible to find 
another species that better agrees with it. It is a very fine new 
species of Phacelophrynium, differing from the generic diagnosis 
in its 1- or 2-celled nuts (instead of 3-celled 3-valved capsules) 
and its long corolla-tube. There is, however, only one exterior 
staminode, and this character together with the whole habit 
of the species, which closely resembles Phacelophrynium inter- 
ruptum K. Schum., induces me to place it in Phacelophrynium 
rather than to propose for it a new generic name. The fruit 
is rather remarkable and is to be compared only with that of 
Halopegia K. Schumann. 

COM IN SI A Hemsley 

COMINSIA GIGANTEA (Scheff.) K. Sch. in Engl. Pflanzenreich 11 (1902) 
58, /. 10. 

Phrynium giganteum Scheff. in Ann. Jard. Bot. Buitenz. 1 (1876) 58. 
Folium buccinatum asperum Rumph. Herb. Amb. 5: 142, t. 62, f. 1 
(sphalm. 2). 

Not represented in Robinson's Amboina collections, but col- 



168 



RUMPHIUS'S HERBARIUM AMBOINENSE 



lected in Amboina by Boerlage and known from Halmaheira, 
New Guinea, Bismarck Archipelago, and the Solomon Islands. 
Rumphius's description and figure unmistakably refer to Comin- 
sia gigantea K. Sch. Burman f . erred in referring fig. 2, of plate 
62 to Folium buccinatum asperum Rumph. in the explanation of 
the plate; it is Folium buccinatum ~Rumi)h.=Heliconia bihai 
Linn., as the description shows. Horaninow thought it might 
be a Phrynium, and Teysmann, quoted by Hasskarl, thought that 
it might be a Hellenia (=Alpinia) ; see Heliconia bihai Linn., 
page 150. 

COMINSIA RUBRA Val. sp. nor. 

Folium mensarium rubrum s. latifolium Rumph. Herb. Amb. 5: 141. 

Herba multicaulis, caulibus complanatis e vaginis 4 com- 
positis. Folia magna late elliptica (Heliconia Bihai referentia), 
antice rotundata acumine convoluto, valde inaequilatera, 600 ad 
900 mm longa 180 ad 250 mm lata consistentia firma, haud ad 
latera findentia, supra in vivo glauco-viridia subtus pallide fusca 
vel violacea. Nervi laterales circ. 12 ad 15 mm inter se remoti. 
Petioli ad 750 mm longi basi cum ligula pubescentes. Herba 
florens ad 2.5 metr. alta, 15 mm crassa, pilis bulbillosis densis 
scabra, foliis 2, 1,200 mm longis 400 mm latis. Panicula magna 
terminalis ad 400 mm longa ramosissima rachi laxe pubescente, 
valde densiflora, internodiis denudatis, inferioribus ad 45 mm, 
superioribus ±10 mm longis. Bracteae vaginantes 35 ad 45 
mm longae glabrae. Flores ignoti. 

The description is from specimens in the Buitenzorg herb- 
arium, collected in Amboina by Botter, and bearing the native 
name kokin merah, which is also cited by Rumphius. It is 
certainly a species of Cominsia, well characterized by the scabrid 
sheaths. It is distinguished by the much denser, pubescent 
inflorescence and by the wider leaves, which are purple beneath. 

ORCHIDACEAE * 
(By J. J. Smith) 
PLATANTHERA L. C. Richard 
PLATANTHERA SUSAN NAE (Linn.) Lindl. Gen. et Sp. Orch. (1835) 295. 
Orchis susannae Linn. Sp. PI. (1753) 939. 

* All specimens of Orchidaceae collected by Doctor Robinson in Amboina 
were sent me at Leiden. Owing to the abnormal conditions brought about by 
the European war, I considered it inadvisable to take the specimens with 
me on my return to Buitenzorg, so that it has been impracticable to cite 
the numbers collected by Doctor Robinson in this consideration of the Orchid- 
aceae described by Rumphius. [J. J. S.]. For citations of Robinson's 
specimens of Orchidaceae, see Addenda, p. 548. [E. D. M.] 



ORCHIDACEAE 169 

Orchis gigantea Sm. Exot. Bot. 2 (1804-05) 79, t. 100. 
Habenaria gigantea Don Prodr. (1825) 24. 
Habenaria susannae R. Br. Prodr. (1810) 312. 
Platanthera gigantea Lindl. in Wall. Cat. (1832) no. 7052. 
Flos susannae Rumph. Herb. Amb. 5: 286, t. 99. 

I have no doubt that Flos susannae minor Rumph. belongs 
with Habenaria rumphii Lindl. 

PERISTYLUS Blume 

PERISTYLUS sp. 

Orchis amboinica minor altera Rumph. Herb. Amb. 6: 118, t. 44, /• 3. 

The plant figured by Rumphius is certainly a Peristylus. I 
do not know any species like it from Amboina. 

HABENARIA Willdenow 

HABENARIA RUMPHII (Brongn.) Lindl. Gen. et Sp. Orch. (1835) 320. 
Platanthera rumphii Brongn. Bot. Duperrey Voy. (1829) 104, t. 38, 
f. A. 

Orchis amboinica minor Rumph. Herb. Amb. 5: 287; 6: 118, t. 54, f. 2. 
Flos susannae minor Rumph. Herb. Amb. 5: 287. 

ANOECTOCHILUS Blume 

ANOECTOCHILUS REIN WARDTII Blume PI, Jav. Orch. (1858) 48, t. 
12, f. 2; t. 12b, f. U; J. J. Sm. Orch. Amb. (1905) 12. 
Folium petolatum Rumph. Herb. Amb. 6: 93. 

Folium petolatum femina s. vera Rumph. Herb. Amb. 6: 92, t. hi, /. 3. 

The flowers of this plant are unknown, so that the determina- 
tion is somewhat doubtful. It is certainly not Macodes petola 
Lindl. 

ZEUXINE Lindley 

ZEUXINE AM BOI N ENSIS J. J. Sm. in Ic. Bog. 2 (1905) 259; Schltr. in 
Bull. Herb. Boiss. II 6 (1906) 298. 
Haplochilus amboinense J. J. Sm. in Bull. Inst. Bot. Buitenz. 7 
(1900) 2. 

Folium petolatum mas Rumph. Herb. Amb. 6: 92, t. 41, f. 2? 

This species was not described by Rumphius. The figure, 
which presents a sterile plant, doubtless represents a species of 
the Physurinae and may belong to Zeuxine ambomensis J. J. Sm. 

COELOGYNE Lindley 

COELOGYNE RUMPHII Lindl. Fol. Orch. Coelogyne (1854) 14 (type!); 
J. J. Sm. Orch. Amb. (1905) 16. 

Pleione rumphii O. Kuntze Rev. Gen. PI. (1891) 690. 

Epidendrum nervosum Lam. Encycl. 1 (1783) 186 (type!), non Coe- 
logyne nervosa Rich. 

Coelogyne psittacina Reichb. f. Xen. Orch. 2: 141, t. 153". 

Angraecum nervosum Rumph. Herb. Amb. 6: 106, t. 48. 



170 



RUMPHIUS'S HERBARIUM AMBOINENSE 



CALANTHE* R. Brown 

CALANTHE VERATRI FOLI A R. Br. in Bot. Reg. 9 (1823) t. 270. 
Limodorum veratri folium Willd. Sp. PL 4 (1805) 122. 
Calanthe furcata Batem. in Bot. Reg. 24 (1838) Misc. 28. 
Ambly glottis veratrifolia Blume Bijdr. (1825) 370. 
Limodorum ventricosum Steud. Nomencl. ed. 1 (1821) 481. 
Calanthe triplicata Ames in Philip. Journ. Sci. 2 (1907) Bot. 326, non 

Orchis triplicata Willem. 
Flos triplicatus Rumph. Herb. Amb. 6: 115, t. 52, f. 2. 

Mr. Oakes Ames f has adopted the name Calanthe triplicata 
(Willem.) Ames for the plant usually known as Calanthe verat- 
rifolia R. Br. on the supposition that Orchis triplicata Willem. 
was the oldest name for the species. From the original descrip- 
tion of Orchis triplicata Willem., for a copy of which I am 
indebted to Mr. Merrill, this was certainly not based on Flos 
triplicatus Rumph., although the Rumphian name was cited as 
a synonym, but was based on a Mauritius plant, presumably 
Calanthe sylvatica LindL, so that Calanthe triplicata Ames, not 
Orchis triplicata Willem., becomes a synonym of C. veratrifolia 
R. Br. Willdenow's description of Limodorum veratrifolium is 
too vague to determine whether it was intended for Calanthe 
veratrifolia R. Br. or C. sylvatica Lindl. He cites both Flos 
triplicatus Rumph. and Orchis triplicata Willem. as synonyms 
and gives the distribution as "Ind. Or." Even if Limodorum 
veratrifolium Willd. belongs with Calanthe sylvatica LindL, 
both Calanthe veratrifolia R. Br. and Amblyglottis veratrifolia 
Blume go entirely with the Malayan plant. It is generally 
accepted that Calanthe veratrifolia was published by Roxburgh 
in Bot. Reg. 7 (1821) sub. t. 573; this is incorrect, as Roxburgh 
merely states that the genus Calanthe should be separated from 
Limodorum, but makes no new combination. Calanthe veratri- 
folia appears first in Bot. Reg. 9 (1823) t. 270. The form 
described by Rumphius may eventually prove to belong to a 
different species. 

PHAJUS Loureiro 

PHAJUS AM BOI N ENSIS Blume Mus. Bot. 2 (1856) 180; J. J. Sm. Orch. 

Amb. (1905) 21. 

Phajus zollingeri Rchb. f. Xenia Orch. 2: 201, Bonplandia 5 (1857) 42. 
Angraecum terrestre alterum Rumph. Herb. Amb. 6: 113, t. 52, /. 1, 
non t. 50, /. 3. 

In the Herbarium Amboinense t. 52, f. 1, is erroneously 

* Retained name, Vienna Code; Alismorkis Thou. = Alismorchis Thou. 
(1809) is older. 

f Philip. Journ. Sci. 2 (1907) Bot. 326. 



ORCHIDACEAE 



171 



referred to Angraecum terrestre primum purpureum (Spatho- 
glottis plicata BL), which was not figured; and t. 50, f. 3, to 
Angraecum terrestre alterum (Phajus amboinensis BL), with 
which t. 52, /. 1, doubtless belongs. Linnaeus and Blume did 
not notice these errors, and therefore the two artificial species 
Epidendrum terrestre Linn, and Phajus rumphii Blume, both 
composed of the characteristics of the two species mentioned 
above, originate from this confusion. Linnaeus's original de- 
scription of Epidendrum terrestre, which was not available in 
Buitenzorg, was copied and sent to me by Mr. Merrill. It is 
as follows : 

Epidendrum terrestre. E. fol. radicalibus lanceolatis nervosis membra- 
naceis, scapo vaginato, petalis oblongis, nectario cymbiformi bifido. Rumph. 
amb. 6, t. 52, f. 1 [Linnaeus Syst. ed 10 (1759) 1246]. 

Apart from the characteristics Angraecum terrestre primum 
purpureum Rumph. and Angraecum terrestre alterum Rumph. 
have in common, that is "foliis lanceolatis nervosis membra- 
naceis, scapo vaginato," the "folia radicalia" occur only in 
Spatho glottis, while in Angraecum terrestre alterum Rumph. 
the leaves are placed on an elongated erect stem. The nectary 
(lip) is cymbiform only in Angraecum terestre alterum Rumph. 
(Herb. Amb. 6: 114), and bifid only in Angraecum terrestre 
primum purpureum Rumph. (Herb. Amb. 6: 112). 

Linnaeus, Sp. PL ed. 2 (1763) 1352, described Epidendrum 
tuberosum as follows: 

Epidendrum foliis lato-lanceolatis nervosis membranaceis bulbo innatis, 
scapo vaginato, nectario cymbiformi bifido. 

Helleborine purpurea, tuberosa radice. Plum. spec. 9. ic. 186. f. 2. 
Angraecum terrestre primum. Rumph. amb. 6. p. 112 t. 52. f. 1. 
Habitat in Indiis. 

This too is a species of doubtful status, but I suppose that it 
has the same value as Epidendrum terrestre Linn. It is cer- 
tainly not Helleborine purpurea, tuberosa radice Plum., for 
Plumier's species does not have a bifid nectary, nor can it be 
Angraecum terrestre primum Rumph., because the nectary of 
Rumphius's species is not cymbiform. Further it cannot be 
Angraecum terrestre alterum Rumph. as this has no bulbs and 
the nectary is not bifid. 

PHAJUS GRATUS Blume Mus. Bot. 2 (1856) 181. 
Limatodes grata Miq. Fl. Ind. Bat. 3 (1855) 672. 

Angraecum terrestre primum album Rumph. Herb. Amb. 6: 113, 
t 50, f. 3. 

This species is unknown to me. From Rumphius's description 



172 



RUMPHIUS'S HERBARIUM AMBOINENSE 



one would think that it referred to a white-flowered form of 
Spatho glottis, and, if this suggestion is correct, Blume's descrip- 
tion applies only to the species figured by Rumphius, t. 50, f. 3. 
It is to be noted that Blume gives a detailed description of the 
lip of Phajus gratus BL, although he certainly did not see a 
specimen, and Rumphius does not describe the lip. 

SPATHOGLOTTIS Blume 

SPATHOGLOTTIS PLICATA Blume Bijdr. (1825) 401, Tabell. /. 76; J. 
J. Sm. Orch. Amb. (1905) 24. 
Angraecum terrestre primum purpureum Rumph. Herb. Amb. 6: 112. 

Rumphius's description applies to Spatho glottis plicata Blume. 
In the Herbarium Amboinense t 52, /. 1 is erroneously referred 
here; it represents Phajus amboinensis Blume. 

Phajus rumphii Blume, Mus. Bot. 2 (1856) 179, is an artificial 
species based on the characters of Angraecum terrestre primum 
purpureum Rumph. {Spatho glottis plicata Blume) and An- 
graecum terrestre alterum Rumph. (Phajus amboinensis 
Blume). 

EULOPHIA * R. Brown 

? EULOPHIA sp. 

Orchis amboinica major radice digitata Rumph. Herb. Amb. 6: 116. 

Hasskarl suggests that this plant may be Cyrtoptera ensifor- 
mis Lindl. Rumphius's description applies closely to Eulophia, 
but I do not know any species of the genus to which the state- 
ment "radix digitata" is applicable. 

LI PAR IS L. C. Richard 

LIPARIS TREUBII J. J. Sm. nom. nov. 

Liparis amboinensis J. J. Sm. in Bull. Jard. Bot. Buitenz. 13 (1914) 6, 

non Orch. Amb. (1905) 31. 
Liparis confusa J. J. Sm. var. amboinensis J. J. Sm. Orch. Amb. 
(1905) 35. 

Angraecum gajang Rumph. Herb. Amb. 6: 108. 

There is no doubt that Angraecum gajang Rumph. is a Liparis. 
A careful study of the question of its identity has convinced me 
that it is the form I described as Liparis amboinensis, here 
called L. treubii; Rumphius's description agrees entirely with 
mine. In raising Liparis confusa var. amboinensis to specific 
rank, I unfortunately overlooked the fact that the name was 
preoccupied in the genus. 

* Retained name, Vienna Code; Graptorkis Thou. = Graptorchis Thou. 
(1809) is older. 



ORCHIDACEAE 



173 



DENDROBIUM * Swartz 

DENDROBIUM CALCEOLUM Roxb. Hort. Beng. (1814) 63, nomen nudum, 
Fl. Ind. ed. 2, 3 (1832) 488. 
Dendrobium roxburghii Lindl. in Journ. Linn. Soc. Bot. 3 (1859) 4. 
Aporum roxburghii Griff, in Calc. Journ. Nat. Hist. 5 (1835) 370. 
Herba supplex quinta Rumph. Herb. Amb. 6: 111, t. 51, f. 2. 

The citation of Rumphius by Roxburgh is incomplete, and as 
to the figure is erroneous. 

DENDROBIUM AC 1 N AC I FORM E Roxb. Hort. Beng. (1814) 63, nomen 
nudum, Fl. Ind. ed. 2, 3 (1832) 487. 
Dendrobium scalpelliforme T. & B. in Nat. Tijdschr. Ned. Ind. 27 

(1864) 17; J. J. Sm. Orch. Amb. (1905) 111. 
Herba supplex major prima s. Herba supplex femina Rumph. Herb. 
Amb. 6: 111. 

Dendrobium acinaciforme Roxb. and D. scalpelliforme T. & B. 
are undoubtedly synonymous, and I think that the reduction of 
Herba supplex major prima to Dendrobium acinaciforme Roxb. 
is the correct disposition of it. The yellow color of the flowers is 
a character so uncommon in the group that it serves as a valuable 
indication of the identity of the Rumphian species. However, 
I have never seen specimens with stems as long as those de- 
scribed by Rumphius. It is certainly not Dendrobium calceolum 
Roxb., which is Herba supplex quinta. Rumphius's t. 51, f. 1, 
can scarcely belong with the plant described as Herba supplex 
prima, as he states that the leaves are in shape similar to those 
of Herba supplex minor, while the figure presents a species with 
flat, not laterally compressed, leaves ; I do not recognize it. 

DENDROBIUM PAPI LION I FERU M J. J. Sm. Orch. Amb. (1905) 42. 
Dendrobium crumenatum Sw., fl. lilac. Miq. Choix. t. 22, f. 1. 
Angraecum crumenatum Rumph. Herb. Amb. 6: t. U7 , f. 2. 

The figure of Angraecum crumenatum does not represent the 
species generally regarded as Dendrobium crumenatum Sw., but 
the form with purple flowers figured by Miquel and described 
by me as Dendrobium papilioniferum. I have not seen the 
original description of Dendrobium crumenatum Sw., but the 
common Malayan species currently known as D. crumenatum is 
certainly specifically distinct from Dendrobium papilioniferum 
J. J. Sm. 

In the Herbarium Amboinense, explanation of 1. 1*7, it is stated 
that Angraecum crumenatum is described in chapter 57 of the 
Auctuarium. The only plant described in this chapter that 

* Retained name, Vienna Code; Callista Lour. (1790) is older. 



174 



RUMPHIUS'S HERBARIUM AMBOINENSE 



agrees more or less with Dendrobium crumenatum is Angraecum 
angustis crumenis. I have thought that the material Rumphius 
described might have presented two species growing together, 
but after a careful study of the question I am now convinced 
that the plant in question must be an Eria, probably Eria mo- 
luccana Schltr. & J. J. Sm. 

DENDROBIUM EPHEMERUM J. J. Sm. comb. nov. 

Dendrobium papilioniferum J. J. Sm. var. ephemerum J. J. Sm. Orch. 

Amb. (1905) 45. 
Angraecum album minus Rumph. Herb. Amb. 6: 99, t. UU, /• 1. 

The plant is not a variety of Dendrobium papilioniferum J. 
J. Sm. as I formerly supposed, although it is not impossible that 
it is a hybrid. The suggestions of Linnaeus, Willdenow, and 
Hasskarl, who respectively reduced this form to Epidendrum 
spathulatum Linn., Cymbidium ovatum Willd., and Dendrobium 
bursigerum Lindl., are certainly incorrect. The other form de- 
scribed by Rumphius in this chapter is probably another species 
of Dendrobium, certainly not Hysteria veratrifolia Reinw., as 
Hasskarl has suggested. 

DENDROBIUM RUMPHIANUM T. & B. in Nat. Tijdschr. Ned. Ind. 24 
(1862) 317; J. J. Sm. Orch. Amb. (1905) 57. 
Angraecum flavum sextum moschatum s. odoratum Rumph. Herb. 
Amb. 6: 102. 

? Angraecum flavum nonum Rumph. Herb. Amb. 6: 104, excl. fig. 

It is possible that Dendrobium minax Rchb. f ., which I 
formerly reduced to D. rumphianum T. & B., may represent some 
other species. It is very doubtful whether or not the Bali plant 
mentioned by Rumphius belongs with this species. Rumphius's 
description of Angraecum flavum nonum agrees very well with 
this species, but the figure closely resembles that of Angraecum 
flavum septimum. 

DENDROBIUM M I RBELI AN U M Gaudich. Bot. Freyc. Voy. (1826) 423, 

t. 88; J. J. Sm. Orch. Amb. (1905) 56. 
Dendrobium prionochilum Kranzl. in Osterr. Bot. Zeitschr. 44 (1894) 

261 (ex Kranzl.). 
Dendrobium rosenbergii T. & B. in Nat. Tijdschr. Ned. Ind. 24 (1862) 

317. 

Angraecum flavum septimum Rumph. Herb. Amb. 6: 103, t. £5; (t. 

46, f. 2?). 

I cannot distinguish Rumphius's figure of Angraecum flavum 
nonum (t. 46, f. 2) from that of Angraecum flavum septi- 
mum (t. 45). The description of the former, however, agrees 
very well with Angraecum flavum sextum moschatum Rumph. 
=Dendrobium rumphianum T. & B. 



ORCHIDACEAE 



175 



DEN DROBI U M MOLUCCENSE J. J. Sm. in Bull. Jard. Bot. Buitenz. 13 
(1914) 11. 

Dendrobium atropurpureum J. J. Sm. (nec Miq.) Orch. Amb. (1905) 
54. 

Herba supplex minor Rumph. Herb. Amb. 6: 110, t. 50, f. 2. 

The species of the section Oxystophyllum have been frequently 
misunderstood and misinterpreted, partly due to the incomplete 
descriptions. Herba supplex minor is neither Dendrobium con- 
cinnum Miq. nor D. atropurpureum Miq. 

DENDROBIUM PURPUREUM Roxb. Fl. Ind. ed. 2, 3 (1832) 484. 

Dendrobium viridiroseum Rchb. f. in Bonplandia 3 (1855) 226. 
Angraecum purpureum silvestre Rumph. Herb. Amb. 6: 109. t. 50, /. 1. 

There is no doubt as to the correctness of this reduction, which 
was made by Roxburgh. 

DENDROBIUM ANOSMUM Lindl. in Bot. Reg. 21 (1844) Misc. 41. 

Dendrobium superbum Rchb. f. var. anosmum Rchb. f. in Walp. Ann. 
6 (1861) 283. 

Dendrobium superbum Rchb. f. in Walp. Ann. 6 (1861) 282. 
Dendrobium macrophyllum Lindl. Bot. Reg. 25 (1839) Misc. 36, non 
A. Rich. 

Dendrobium macranthum Hook, in Curtis's Bot. Mag. 69 (1843) t. 

3870, non A. Rich. 
Dendrobium scortechinii Hook. f. Fl. Brit. Ind. 5 (1890) 741. 
Dendrobium leucorhodum Schltr. Orch. Neu-Guinea (1913) 499. 
Angraecum caninum s. undecimum Rumph. Herb. Amb. 6: 105, t. 

U7, /. 1. 

I have seen no specimens of Dendrobium anosmum Lindl., but 
as the plant is usually considered to be a mere variety of Dendro- 
bium superbum Rchb. f. there is no good reason for replacing 
Lindley's name by the more recent one proposed by Reichen- 
bach f. The figure of Dendrobium anosmum Lindl., in Paxt. 
Mag. Bot. 15: 97, represents a short-flowered form of Dendro- 
bium superbum Rchb. f . Dendrobium anosmum Lindl. is credited 
to the Philippine Islands by Lindley, Bot. Reg. 31 (1845) Misc. 
32, but is not mentioned by Mr. Ames. 

DENDROBIUM sp. 

Herba supplex major quarta Rumph. Herb. Amb. 6: 111. 

The description of the flower is strongly suggestive of Dendro- 
bium confusum Schltr., but so far as I know Schlechter's species 
never attains the length noted by Rumphius. Suggested reduc- 
tions by other authors are in contradiction with the description. 

DENDROBIUM sp. 

Angraecum purpureum et nudum Rumph. Herb. Amb. 6: 109, t.Jf9,f.2. 

Lindley reduced this form to Dendrobium bifarium Lindl., the 



176 



RUMPHIUS'S HERBARIUM AMBOINENSE 



correctness of which I doubted in my former consideration of the 
Amboina Orchidaceae.* Dendrobium bif avium Lindl. was based 
on a flowerless specimen from Penang and is now recognized as 
a species of the section Distichophyllum with solitary flowers. 
Accordingly Angraecum purpureum et nudum Rumph. cannot 
possibly belong here. The form described by Rumphius is 
apparently a species of Dendrobium, perhaps of the section 
Pedilonum. 

DENDROBIUM sp. 

Angraecum jamboe Rumph. Herb. Amb. 6: 108. 

In my former treatment of the Orchidaceae of Amboina I 
reduced this to Dendrobium pruinosum T. & B., but I now am 
of the opinion that that was an erroneous disposition of it. The 
leaves are not acute, and the lip is not violet-blue. Doctor Rob- 
inson in his field notes suggests that the plant may be Pseuderia 
foliosa Schltr., but there are so many discrepancies between 
Rumphius's description and the characters of Brongniart's spe- 
cies that it is very doubtful if this suggested disposition of it 
is the correct one. I suppose that the plant in question is a 
species of Dendrobium of the section Grastidium. 

? DENDROBIUM sp. 

Herba supplex major tertia Rumph. Herb. Amb. 6: 111. 

The status of this plant is doubtful, but it is probably a species 
of Dendrobium of the section Rhoplalanthe. It certainly cannot 
be referred to Dendrobium atropurpureum Miq. as Hasskarl 
suggested. 

? DENDROBIUM sp. 

Herba supplex major secunda Rumph. Herb. Amb. 6: 111. 

The description is so vague that it is not even certain that the 
plant described by Rumphius belongs in the genus Dendrobium. 
According to the explanation of t. 51, f. 1, this figure represents 

Herba supplex femina s. secunda, not Herba supplex major 
secunda. 

ERIAf Lindley 

ERIA MOLUCCANA Schltr. & J. J. Sm. Orch. Amb. (1905) 74. 
Angraecum angustis crumenis Rumph. Herb. Amb. 6: 107. 

Under Dendrobium papilioniferum J. J. Sm. I have already 
pointed out that this is probably the correct disposition of the 
plant Rumphius described. The data given by Rumphius con- 



* Orch. Amb. (1905) 62, 119. 

f Retained name, Vienna Code; Pinalia Buch.-Ham. (1825) is older. 



ORC H ID ACE AE 



177 



forms closely with the characters of Eria moluccana Schltr. & 
J. J. Sm., the two lateral leafy stems being the inflorescences 
with bracts. 

ERIA sp. 

Angraecum lanuginosum Rumph. Herb. Amb. 6: 108. 

The form described is doubtless an Eria of the section Tricho- 
tosia. No species of this section has as yet been described from 
Amboina under the binomial system. 

BULBOPHYLLUM * Thouars 

BULBOPHYLLUM sp. 

Angraecum uniflorum Rumph. Herb. Amb. 6: 107. 

The description of this plant in almost every point applies 
to Bulbophyllum grandiflorum Blume, but so far as I know 
Blume's species has not been recorded from Amboina. 

GRAMMATOPHYLLUM Blume 

GRAM M ATOPH YLLU M SCRIPTUM (Linn.) Blume Rumphia 4 (1848) 
48; J. J. Sm. Orch. Amb. (1905) 84. 
Grammatophyllum rumphianum Miq. Ann. Mus. Bot. Lugd. Bat. 4 

(1869) 219, t. 8, 9. 
Grammatophyllum speciosum Lindl. Gen. et. Sp. Orch. (1833) 173, p. p. 
Grammatophyllum leopardinum Rchb. f. in Flora 46 (1888) 151. 
Grammatophyllum guilelmi II Kranzl. in Gartenfl. 43 (1894) 114. 
Gabertia scripta Gaudich. Bot. Freyc. Voy. (1826) 425. 
Vanda scripta Spreng. Syst. 3 (1826) 719. 
Cymbidium scriptum Sw. in Schrad. Journ. (1799) 218. 
Epidendrum scriptum Linn. Sp. PL ed. 2 (1763) 1351 (type!). 
Angraecum scriptum Rumph. Herb. Amb. 6: 95, t. Jf2. 

In addition to the above form Rumphius describes two other 
species in this chapter, one growing on Mangifera and other 
trees, the other on Cocos. Hasskarl suggests that the first may 
be Cymbidium wallichii Lindl., while Blume refers the latter, 
from its Portuguese name Fulha alacra, to Arachnis flos aeris 
Rchb. f. The suggested reductions of both, however, are con- 
tradicted by Rumphius's descriptions. 

PHALAENOPSIS Blume 

PHALAENOPSIS AMABILIS (Linn.) Blume Bij dr. (1825) 294, Tabell. /. U. 

Epidendrum amabile Linn. Sp. PI. (1753) 953. 
Cymbidium amabile Roxb. Hort. Beng. (1814) 63. 
Phalaenopsis grandiflora Lindl. in Gard. Chron. (1848) 39. 
Angraecum album majus Rumph. Herb. Amb. 6: 99, t. US. 

Rumphius briefly describes two forms which, without doubt, 

♦Retained name, Vienna Code; Phyllorkis Thou. = Phyllorchis Thou. 
(1809) is older. 

144971 12 



178 



RUMPHIUS'S HERBARIUM AMBOINENSE 



are referable to the same species. It is well known that the 
flowers of Phalaenopsis amabilis Blume vary considerably in 
size, in the form of the sepals, petals, lip, and especially in the 
size and markings of the yellow area on the lip. Specimens 
with the sepals purplish on the outside are not rare. Rumphius's 
description of the first form does not at all apply to Phalaenopsis 
violacea T. & B., to which it was reduced by Has§karl. 

LUISIA Gaudichaud 
LUISIA CONFUSA Rchb. f. in Walp. Ann. 6 (1861) 621. 

Luisia teretifolia Blume (non Gaudich.) Rumphia 4 (1848) t. 19U, f- 
3, t. 197D. 

Angraecum decimum et angustifolium Rumph. Herb. Amb. 6: 104. 

Luisia teretifolia Gaudich., as formerly interpreted, contained 
several distinct species. I think that Reichenbach was correct 
in separating from it the Amboina form under the name Luisia 
confusa Rchb. f . 

VAN DA R. Brown 

VAN DA FURVA Lindl. (non Blume) Gen. et Sp. Orch. (1833) 215; J. 
J. Sm. Orch Amb. (1905) 98. 
Cymbidium furvum Willd. Sp. PI. 4 (1805) 103. 

Epidendrum furvum Linn. Sp. PI. ed. 2 (1763) 1348 (type!), excl. 
syn. Rheede. 

Angraecum octavum et furvum Rumph. Herb. Amb. 6: 104, t. 46, f. 1. 
VAN DA sp. 

Vanda crassiloba T. & B. (non J. J. Sm.) in Cat. Hort. Bogor. (1866) 

48, nomen nudum. 
Angraecum saxatile Rumph. Herb. Amb. 6: 107, t. 49, f. 1. 

The plant I described in my former treatment of the Orchida- 
ceae of Amboina as Vanda crassiloba, is certainly not the same as 
the species Teysmann and Binnendijk had in view. The latter 
is very similar to Vanda celebica Rolfe and will probably prove 
to be a variety of this or a closely allied species when fresh 
material from Amboina shall have become available. The spe- 
cies of the Vanda hastifera group are in want of revision. It is 
evidently quite different from Cymbidium, where it was placed 
by Hasskarl. 

VANDOPSIS Pfitzer 

VANDOPSIS LISSOCHILOIDES (Gaudich.) Pfitz. in Engl. & Prantl Nat. 

Pflanzenfam. 2 6 (1889) 210, /. 229. 
Fieldia lissochiloides Gaudich. Bot. Freyc. Voy. (1826) 424, t. 36. 
Vanda lissochiloides Lindl. Gen. et Sp. Orch. (1833) 216. 
Vanda batemannii Lindl. in Bot. Reg. (1846) t. 59. 
Stauropsis lissochiloides Benth. ex Pfitz. Vergl. Morph. Orch. (1882) 

14. 

Angraecum quintum Rumph. Herb. Amb. 6: 102. 



CASUARINACEAE 



179 



RENANTHERA Loureiro 

RENANTHERA MOLUCCANA Blume Rumphia 4 (1848) 54, t. 193, f. 
2; t. 197E. 

Angraecum rubrum Rumph. Herb. Amb. 6: 101, t. UU, f- 2. 

I do not understand the exact status of the form of this plant 
described by Rumphius. 

SARCANTHUS Lindley 

SARCANTHUS SUBULATUS (Blume) Reichb. f. in Bonplandia 5 (1857) 
41; J. J. Sm. Orch. Amb. (1905) 103. 
Sarcanthus secundus Griff. Not. 3 (1851) 362. 
Cleisostoma subulatum Blume Bijdr. (1825) 363. 
Angraecum pungens Rumph. Herb. Amb. 6: 106. 

I have little doubt that the plant described by Rumphius as 
Angraecum pungens is the same as Sarcanthus subulatus Rchb. 
f. Usually the sepals and petals are light or dark brown with 
a pale center, but the specimen I collected in Amboina had very 
pale flowers. Schoenorchis juncifolia Blume, to which Hasskarl 
reduced it, is a totally different plant. 

ORCHIDACEAE OF UNCERTAIN STATUS 

Angraecum sediforme Rumph. Herb. Amb. 6: 107. 

Unrecognizable. • 
Angraecum taeniosum Rumph. Herb. Amb. 6: 108. 

The plant described by Rumphius evidently belongs in the Sarcanthinae. 
Maccabuhay Rumph. Herb. Amb. 5: 287. 

It is not at all certain that this plant is an orchid. The material was 
from the Philippines; see Tinospora, Menispermaceae. 
Satyria Rumph. Herb. Amb. 5: 287. 

Unrecognizable. Rumphius's material was from China, the Chinese 
name pu sang tjan being quoted by him. 

ANGIOSPERMAE 

(DICOTYLEDONS) 

CASUARINACEAE 

CASUARI N A Linnaeus 

CASUARI N A RUMPHIANA Miq. in Flora 48 (1865) 23 (type!), 38. 
Casuarina montana Rumph. Herb. Amb. 3: 87, t. 58 (excl. /. A). 
Amboina, Soja, Robinson PI. Rumph. Amb. 411, October 24, 1913, alti- 
tude about 275 meters, locally known as kasuari. 

Casuarina montana is the whole basis of Casuarina rumphiana 
Miq., as originally published in Flora 48 (1865) 23, but on page 
38 of the same volume Miquel amplifies the description from 



180 



RUMPHIUS'S HERBARIUM AMBOINENSE 



specimens collected in Amboina by Teysmann and DeVriese. 
In his Revisio critica Casuarinarum (1848) 282 and in his Flora 
Indiae Bataviae l 1 (1858) 873, Miquel erroneously referred Ca- 
suarina montana Rumph. to Casuarina nodiflora Forst. The 
species is a characteristic one of the mountain forests of the 
Philippines and of the Moluccas. 

CASUARINA EQUISETI FOLIA Linn. Amoen. Acad. 4 (1759) 143 (type!) 
(equisefolia) ; Forst. Char. Gen. (1776) 103. t. 52. 
Casuarina litorea Rumph. Herb. Amb. 3: 86, t. 57. 

This widely distributed and well-known species is not repre- 
sented in our Amboina collections. Rumphius's figure is an 
excellent one, and there is no doubt whatever as to the correctness 
of this reduction, which was originally made by Linnaeus; in 
fact Casuarina litorea Rumph. is the type of Casuarina equiseti- 
folia Linn, as published in Amoen. Acad. 4 (1759) 143, although 
the species is usually accredited to Forster (1^76). 

CASUARINA SUMATRANA Jungh. in Hoev. & DeVriese Tijdschr. 11 
(1844) 115. 

Casuarina celebica Rumph. Herb. Amb. 3: 87, t. 58 f. A. 

This disposition of Casuarina celebica follows Miquel and is 
certainly correct. The species is well characterized by its re- 
latively large fruits. 

PIPERACEAE * 
PIPER Linnaeus 

PIPER ARBORESCENS Roxb. Hort. Beng. (1814) 80 (type!), Fl. Ind. 
1 (1820) 161, ed. 2, 1 (1832) 159. 

Piper miniatum Blume in Verh. Bat. Genoots. 11 (1826) 166. 
Sirium arborescens tertium Rumph. Herb. Amb. 5: 46, t. 28, f. 1. 

Amboina, Soja, Way tommo, Koeda mati, Gelala, Amahoesoe, and Halong, 
Robinson PI. Rumph. Amb. 62, 63, 6U, 609, August and September, 1913, 
locally known siri seytan and siri tulak tulak. 

The Rumphian illustration is the whole basis of Piper arbores- 
cens Roxb. as originally published in the Hortus Bengalensis 

* I am under obligations to Mr. C. de Candolle, Geneva, Switzerland, 
for determinations of the Piperaceae collected by Doctor Robinson. In 
all but two cases the specimens are here quoted under the binomials indi- 
cated by him, the exceptions being Piper decumanum Linn., reported as 
P. forstenii C. DC., and Piper arborescens Roxb., reported as P. miniatum 
Blume, the specific name adopted by me being accepted as the oldest valid 
one in each case. The determinations of the Rumphian species are by 
Doctor Robinson and myself. 



PIPERACEAE 



181 



(1814) 80.* The species was later briefly described by Roxburgh 
from specimens collected in the Moluccas, the reference to 
Rumphius being included in the description. The description 
applies to the above specimens in all respects, and is not at all 
Thottea dependent Klotzsch, as reduced in Index Kewensis. The 
actual Amboina specimens were determined by C. de Candolle 
as Piper miniatum Blume, Blume's species thus becoming a 
synonym of Piper arborescens Roxb. In Stickman's Herb. Amb. 
(1754) 19, it is erroneously referred to Piper malamiris L. 

PIPER CANINUM Blume in Verh. Bat. Genoots. 11 (1826) 214, var. 
GLABRI BRACTEU M C. DC. in Philip. Journ. Sci. 5 (1910) Bot. 
459. 

Piper caninum Rumph. Herb. Amb. 5; 49, t. 28, f. 2. 
Amboina, Wae and Amahoesoe, Robinson PI. Rumph. Amb. 57, August 
and November, 1913, on coral rocks and on trees at low altitudes. 

Blume was certainly correct in referring this Rumphian 
species to Piper caninum Blume. Loureiro, Fl. Coehinch. (1790) 
30, discussed it under Piper sylvestre Lour., while Roxburgh, 
Fl. Ind. 1 (1820) 161, reduced it to Piper cubeba; Piper cubeba 
Roxb., non Linn., is a synonym of Piper caninum Blume. 

PIPER DECUMANUM Linn, in Stickman Herb. Amb. (1754) 19, Amoen. 
Acad. 4 (1759) 128 (type!), Syst. ed. 10 (1759) 856, Sp. PL ed. 2 
(1762) 41, excl. syn. Plumier. 
Piper forstenii C. DC. Prodr. 16 1 (1869) 348. 
Sirium decumanum Rumph. Herb. Amb. 5: 45, t. 27. 
Amboina, Hatiwe, Batoe merah, and Oerimesseng, Robinson PI. Rumph. 
Amb. 61, August and September, 1913, climbing on trees, altitude 10 to 
250 meters. 

Sirium decumanum Rumph. is the whole basis of Piper decu- 
manum Linn, as originally published and is hence the type of 
the species; Linnaeus took his specific name decumanum from 
Rumphius. Later he added also a reference to Plumier, Amer. 
59, t. 76, which represents a species totally different from Sirium 
decumanum. C. de Candolle, Prodr. 16 1 (1869) 370, interpreted 
Piper decumanum Linn, from the reference to Plumier and de- 
scribed the Amboina species as Piper forstenii C. DC. Some of 
the early authors referred it to Piper methysticum Forst., bub 
it has nothing to do with Forster's species. It has al§o been cited 
under Piper majusculum Blume and Chavica majuscula Miq., 
while Miquel referred it to Chavica rumphii Miq. Both Blume's 
and Miquel's species are apparently different from Piper decu- 
manum Linn. (P. forstenii C. DC.) as here interpreted. 

* See C. B. Robinson in Philip. Journ. Sci. 7 (1912) Bot. 415. 



182 



RUMPHIUS'S HERBARIUM AMBOINENSE 



PIPER BETLE Linn. Sp. PI. (1753) 28. 

Siriifolium Rumph. Herb. Amb. 5: 336, t 116, f. 2. 

The common betel pepper is not represented in our Amboina 
collections, yet Piper betle is manifestly the correct disposition 
of Siriifolium Rumph. Linnaeus reduced it, through error, to 
Piper longum Linn., in Stickman Herb. Amb. (1754) 22, Amoen. 
Acad. 4 (1759) 131, and later, Sp. PI. ed. 2 (1762) 41, placed 
it under Piper malamiris Linn., which is apparently a synonym of 
Piper betle Linn. Radermacher, Loureiro, Roemer and Schultes, 
Blume, and other authors cite it under Piper betle Linn., and 
Miquel cites it under Chavica betle Miq. 

PIPER BETLE Linn. var. SIRIBOA (Linn.) C. DC. Prodr. 16 1 (1869) 359. 
Piper siriboa Linn. Sp. PI. (1753) 29. 
Siriboa Rumph. Herb. Amb. 5: 340, t. 117. 

Siriboa Rumph. was originally reduced to Piper siriboa Linn., 
in Stickman Herb. Amb. (1754) 22, one year after the publica- 
tion of the species, and has been very consistently cited under 
this name in subsequent botanical literature. The forms de- 
scribed by Rumphius as / alba, II cambing, and 77/ fragrans 
are apparently but variants of Piper betle or the variety siriboa. 

PIPER AMBOINENSE (Miq.) C. DC. Prodr. 16 2 (1869) 347. 

Chavica amboinensis Miq. Ann. Mus. Bot. Lugd. Bat. 1 (1863—64) 
i . 134. 

Sirium arborescens tertium alterum Rumph. Herb. Amb. 5: 48. 
Amboina, Mahija, Batoe gadjah, and near the town of Amboina, Robin- 
son PL Rumph. Amb. 58, ascending to an altitude of 250 meters, climbing 
on trees, locally known as siri seytan. 

The specimen cited above represents a characteristic species 
of Piper, and it is probably the form described by Rumphius, here 
reduced to Piper amboinense C. DC. 

PIPER REIN WAR DTI AN U M (Miq.) C. DC. Prodr. 16 1 (1869) 354. 

Macropiper reinwardtianum Miq. in Linnaea 21 (1848) 481. 

Sirium decumanum album Rumph. Herb. Amb. 5: 45. 
Ambonia, Wae, Lateri, and Halong, Robinson PL Rumph. Amb. 60, 
August and November, 1913, in light forests, altitude from sea level to 
150 meters, locally known as siri utan, siri tallan, and siri tallan perampuan. 

The specimen cited here probably represents the form that 
Rumphius described. Vahl, Enum. 1 (1804) 334, referred the 
Rumphian species to Piper album Vahl, a species based on Javan 
specimens and one of doubtful status. There is no particular 
reason for believing that the Amboina' plant is the same as the 
Javan one described by Vahl, but it does appear from the descrip- 
tion to be referable to Piper reinwardtianum C. DC. 



PIPERACEAE 



183 



PIPER RETROFR ACTUM Vahl Enum. 1 (1804) 314. 
Piper chaba Hunter in As. Res. 9 (1809) 391. 
Chavica officinarum Miq. Syst. Pip. (1844) 256. 
Piper officinarum C. DC. Prodr. 16 1 (1869) 356. 
Piper longum Rumph. Herb. Amb. 5: 333, t. 116, f. 1. 
Piper longum e Philippinis Rumph. Herb. Amb. 5: 334. 
Pharmacum magnum vulgare Rumph. Herb. Amb. 5: 42, t. 26, f. 2? 

Piper retrofractum Vahl is not represented in our Amboina 
collections. The figure of Piper longum Rumph., however, 
seems to represent Vahl's species, this reduction being in agree- 
ment with Miquel and with C. de Candolle. The form from the 
Philippines I have determined largely from the native name 
cited by Rumphius, sabia being the universal name, at least 
about Manila, for Piper retrofractum Vahl. I follow Miquel 
also in reducing here Pharmacum magnum vulgare Rumph., 
who considered that it represented a form near Chavica officina- 
rum Miq. Linnaeus reduced Piper longum Rumph. to Piper 
amalago Linn*, in Stickman Herb. Amb. (1754) 22, Amoen. 
Acad. 4 (1759) 131, Syst. ed. 10 (1759) 856, Sp. PL ed. 2 (1762) 
41, which was an entirely erroneous disposition of it, as Piper 
amalago Linn, is an American species. Loureiro, Fl. Cochinch. 
(1790) 32, erroneously reduced it to Piper longum Linn. 
Roxburgh, Fl. Ind. 1 (1820) 158, placed it under Piper chaba 
Hunter, which is a synonym of P. retrofractum Vahl. Poiret, in 
Lamarck Encycl. 5 (1804) 460, erroneously placed it under Piper 
plantagineum Lam., and it has been cited under Chavica 
officinarum Miq. and Piper officinarum C. DC, both synonyms of 
Piper retrofractum Vahl. 

PIPER CADUCI BRACTEU M C. DC. sp. nov. 

Sirium silvestre Rumph. Herb. Amb. 5: 342, t. 118, f. 1, 2. 
Amboina, Halong and Koeda mati, Robinson PL Rumph. Amb. 59 (type!), 
August and September, 1913, climbing on trees at low altitudes, locally 
known as siri seytan; Teysmann in Herb. Bogor. 

Ramulis glabris ; f oliis breviter petiolatis glabris, limbo ovato- 
elliptico basi leviter inaequilatera' utrinque acuto apice acute 
acuminato, in mare 5-plinervio, in femine 9-ninervio nervo cen- 
trali nervos adscendentes 2 mittente quorum supremus 2 in 
mare et 3 in femina a basi solutis quorum externi aliis multo 
breviores et tenuiores, petiolo basi ima vaginante; pedunculo 
glabro petiolum aequante vel paulo superante, stirpis masc. 
spica quam limbus pluries breviore, rhachi hirsuta, bracteae 
pelta rotunda glabra centro-pedicellata pedicello hirsuto, stami- 
nibus 2 antheris minutis rotundatis 4-valvatis filamenta fere 
aequantibus; stirpis fern, spica quam limbus pluries breviore, 



184 



RUMPHIUS'S HERBARIUM AMBOINENSE 



rhachi hirsuta, bracteae pelta glabra elliptica centro pedicellata 
decidua pedicello hirsuto, baccis condensis obovatis glabris, stig- 
matibus rotundatis minutis. 

Dioicum. Ramuli spiciferi in mare 1 mm in femina 1.5 ad 2 
mm crassi, collenchyma libriforme in mare in fasciculos discretos 
dispositum in femina subcontinuum, fasciculi intramedullares 
2-seriati, canalis lysigenus centralis periphericique multi. 
Limbo in sicco firme membranacei minute pellucido-punctulati, 
superi usque ad 18 cm longi et 8 cm lati. Petioli usque ad limbi 
latus longius 10 mm, inter limbi latera 1 mm longi. Spica in 
mare subflorens 4 cm longa et 2.25 mm crassa, in femina usque 
ad 4.5 cm longa et 3 mm crassa. Rhachis in mare et in femina 
canalibus lysigenis periphericis munita, bracteae pelta in mare 
1 mm diam., in femina 1 mm longa et 0.75 mm alta, bacca 1.5 
mm longa in siccb atrorubens vel nigra. 

The above description, based on the two specimens cited above, 
has been kindly supplied by Mr. C. de Candolle of Geneva, Swit- 
zerland. The identification of Sirium silvestre Rumph. with 
Piper caducibracteum C. DC. has been made by myself, following 
Doctor Robinson's suggestion that the specimen collected by 
him possibly represented Rumphius's species. Two forms are 
described and figured by Rumphius, but I consider that Sirium 
silvestre II, at least as figured, represents merely a juvenile 
form of Sirium silvestre I. Linnaeus, in Stickman Herb. Amb. 
(1754) 22, Amoen. Acad. 4 (1759) 131, Syst. ed. 10 (1759) 856, 
referred it to Piper malamiris Linn., where it certainly does not 
belong. Miquel thought that both forms described by Rumphius 
might be the same as Chavica malamiris Miq., which is merely 
a synonym of Piper malamiris Linn. C. de Candolle, Prodr. 16 1 
(1869) 361, cites both figures with doubt under Piper sirium 
C. DC, which is essentially a new name for Chavica malamiris 
Miq.; only Indian specimens are cited, and it is entirely im- 
probable that the Amboina plant cited by Rumphius is the same 
as the Indian one described. • 

PIPER NIGRUM Linn. Sp. PI. (1753) 28. 

Piper album et nigrum Rumph. Herb. Amb. 5: 335. 

This reduction follows Miquel, Syst. Piper. (1844) 309, and 
is unquestionably the correct disposition of the plant that 
Rumphius described. 

PIPER SUBPELTATUM Willd. Sp. PL 1 (1798) 166 (type!). 
Lomba Rumph. Herb. Amb. 6: 133, t. 59, f. 1. 

Linnaeus originally reduced Lomba to Piper peltatum Linn., 
an American species to which it does not belong, in Stickman 



PIPERACEAE 



185 



Herb. Amb. (1754) 27, Amoen. Acad. 4 (1759) 135, Sp. PI. 
ed. 2 (1762) 42, in which he was followed by Burman f., Fl. Ind. 
(1768) 15. It is the type of Piper subpeltatum Willd., which 
C. de Candolle considers to be a variety of Piper umbellatum 
Linn., in Donn.-Sm. Enum. 6:39, Philip. Journ. Sci. 5 (1910) 
Bot. 463. Lomba has also been cited by various authors under 
the synonyms Peperomia subpeltata Dietr., Peperidia subpeltata 
Kostel., and Potemorphe subpeltata Miq. 

PIPER SARMENTOSUM Roxb. Fl. Ind. 1 (1820) 162. 

Sirium terrestre Rumph. Herb. Amb. 5: 344, t. 119, f. 1. 

Amboina, Robinson PL Rumph. Amb. 56, July 17, 1913, along roadsides 
in the vicinity of the town of Amboina, July 17, 1913, locally known as 
siri rambang. 

Miquel, Fl. Ind. Bat. I 2 (1858) 446, thought that Sirium 
terrestre might be Chavica sphaerostachya Miq., but C. de Can- 
dolle, Prodr. 16 1 (1869) 389, definitely excludes it as a synonym 
of Miquel's species, and refers it, 1. c. 360, to Piper arcuatum 
Blume. The specimen, however, which certainly represents the 
form that Rumphius described and figured, is not at all Piper 
arcuatum Blume, but is typical Piper sarmentosum Roxb. 

PIPER sp. 

Sirium frigidum rotundifolium Rumph. Herb. Amb. 5: 345, t. 119, /. 2. 

This is not represented in our Amboina collections. Vahl, 
Enum. 1 (1804) 333, referred it to Piper diffusum Vahl, a species 
based on Ceylon specimens and which is apparently the same 
as Piper argyrophyllum Miq.; see Trimen Fl. Ceyl. 3 (1895) 
429. Miquel, Fl. Ind. Bat. I 2 (1858) 441, thought that it might 
be Piper sarmentosum Roxb., but the other form figured on the 
same plate, Sirium terrestre Rumph., is Piper sarmentosum 
Roxb. and is different from Sirium frigidum rotundifolium 
Rumph. The form described as Sirium frigidum latifolium Rumph., 
Herb. Amb. 5 : 345, is indeterminable. Hasskarl, Neue Schlussel 
(1866) 131, thought that it was Piper album Vahl, but there is 
no reason for considering that this reduction is correct. 

PIPER sp. 

Pharmacum magnum parvifolium Rumph. Herb. Amb. 5: 42, t. 26, 
f. 2. 

Nothing in our Amboina collections matches this figure. The 
species is indeterminable from the data at present available. 

PIPER sp. 

Pharmacum magnum marinum Rumph. Herb. Amb. 5: 42. 

Indeterminable from data at present available. 



186 



RUMPHIUS'S HERBARIUM AMBOINENSE 



FAGACEAE 

QUERCUS Linnaeus 

QUERCU6 MOLUCCA Linn. Sp. PI. (1753) 1199 (type!). 
Quercus molucca Rumph. Herb. Amb. 3: 85, t. 56. 

The Rumphian figure and illustration are the whole basis of 
this species, which is one of the few published in the first 
edition of the Species Plantarum with references to the Her- 
barium Amboinense. The species has been recognized in all 
general works, but it is by no means certain that all of the 
botanical material in herbaria under the name Quercus molucca 
Linn, is of the same form as that figured and described by 
Rumphius. It is to be typified by material from the Sula Islands 
from whence Rumphius received his material. 

A second form, very briefly described, having oblong inedible 
fruits, apparently represents a distinct species of Quercus. To 
this Hasskarl, Neue Schliissel (1866) 54, refers t. 56, but this is 
certainly due to a typographical error, as on the preceding page 
the plate is properly cited under Quercus molucca. 

ULMACEAE 

CELT IS Tournefort 

CELTIS PHILIPPENSIS Blanco PI. Filip. (1837) 197. 
Sirifolia Rumph. Herb. Amb. 3: 64, t. 36. 
Sirifolia litorea Rumph. Herb. Amb. 3: 64, t. 37. 

Possibly two species are represented by the forms Rumphius 
described, but they are at least of the same genus. The figure 
representing Sirifolia presents a seedling and a leafy branch 
only, but the one representing Sirifolia litorea, a branch in fruit, 
is a very good representation of Celtis philippensis Blanco. 
Sirifolia was erroneously reduced by Henschel, Vita Rumph. 
(1833) 155, to Piper malamiris Linn., while Hasskarl, Neue 
Schliissel (1866) 51, thought that it might be Cocculus angusti- 
folius Hassk. Teysmann, quoted by Hasskarl, placed it in Sole- 
nostigma=Celtis, where it certainly belongs. Sirifolia litorea 
Rumph., which more certainly represents Celtis philippensis 
Blanco than the preceding, was thought by Hasskarl, Neue 
Schliissel (1866) 51, to be possibly the same as Cocculus lauri- 
folms DC, which is an entirely erroneous disposition of it. 
Hasskarl also quotes Teysmann's opinion that it was a species 
of Solenostigma= Celtis. If it is not Celtis philippensis Blanco, 
which is widely distributed in the Philippines and extends to 
northeastern Australia, it at least represents a very closely allied 
form. 



ULMACEAE 



187 



TREMA Loureiro 

TREMA AM BOI N ENSIS (Willd.) Blume Mus. Bot. 2 (1856) 61 quoad, 
syn., excl. descr. ! 
Celtis amboinensis Willd. Sp. PI. 4 2 (1805) 997. 

Sponia amboinensis Decne. in Nuov. Ann. Mus. Paris 3 (1834) 498. 
Trema virgata Blume Mus. Bot. 2 (1856) 59. 
Sponia virgata Planch, in Ann. Sci. Nat. Ill 10 (1848) 316. 
Cortex piscatorium Rumph. Herb. Amb. 4: 125, t. 61. 

Amboina, Hitoe messen, Robinson PI. Rumph. Amb. 328, November 5, 
1913, in light forests, altitude about 100 meters, locally known as rufut. 

Hasskarl, Neue Schliissel (1866) 84, suggests that Cortex 
piscatorium may be Sponia timorensis Decne., which is the only 
previously suggested reduction of the Rumphian description and 
figure. Sponia timorensis Decne. = Trema timorensis Blume is 
manifestly closely allied to Trema virgata Blume= Trema am- 
boinensis as here interpreted (not of other authors), which 
Lauterbach * confines to Timor Island, with the var. pallida 
(Blume) Lauterb. in Amboina. The Rumphian figure and de- 
scription are manifestly applicable to Trema virgata Blume, the 
figure presenting equilateral leaves which are not cordate at the 
base and lax inflorescences, while the description definitely states 
that the leaves are: "ad factum rugosa, sed non Lanuginosa." 
The actual Amboina specimen cited above, presents a form with 
rather small leaves, but otherwise agreeing very closely with 
the figure. 

As to the synonymy given above, it is to be noted that Sponia 
amboinensis Decne.= Trema amboinensis Blume was based on 
Celtis amboinensis Willd., and Willdenow's description is very 
definitely applicable to Trema virgata Blume, not to Trema am- 
boinensis as currently interpreted. The type was a specimen 
from Amboina, and the leaves are very definitely described as 
"scabriuscula" with no mention of the indumentum so character- 
istic of Trema amboinensis auct., while they are also very de- 
finitely described as equilateral at the base ; in fact this character 
is the one on which the species was primarily distinguished from 
its congeners. It is very evident that Trema amboinensis of all 
modern authors is not the same as Celtis amboinensis Willd. on 
which it was manifestly based, but the type has been consistently 
misinterpreted. Trema amboinensis of modern authors should 
be reduced to Trema orientalis (Linn.) Blume, at least as a 
variety; while Trema virgata Blume, generally recognized as 
a valid species, becomes a synonym of the true Trema amboi- 



* Engl. Bot. Jahrb. 50 (1913) 317. 



188 RUMPHIUS'S HERBARIUM AMBOINENSE 

nensis (Willd.) Blume (Celtis amboinensis Willd.). Doctor 
Lauterbach, in answer to my query regarding the identity of 
Celtis amboinensis Willd., writes from Breslau as follows : "Das 
Originalexamplar von Celtis amboinensis im Willdenow Herbar 
ist von den Herrn Professoren Volkens imd Gilg mit Trema 
orientalis var. amboinensis Lauterb. verglichen worden. Das- 
selbe stimmt mit derselben nicht iiberein, entspricht dagegen 
meiner Trema virgata Bl. var. scabra (Bl.) Ltb." 

MORACEAE 

MORUS Linnaeus 

MORUS ALBA Linn. Sp. PI. (1753) 986. 

Morus indica Linn. Sp. PI. (1753) 986. 
Moms indica Rumph. Herb. Amb. 7: 8, t. 5. 

The form described and figured by Rumphius is the one de- 
scribed by Linnaeus as Morus indica Linn, and was reduced by 
Linnaeus to this species in his Syst. ed. 10 (1759) 1266. This 
disposition of it has been accepted by all authors who have had 
occasion to cite the Rumphian illustration. Morus indica Linn., 
however, does not appear to be specifically distinct from Morus 
alba Linn. 

BROUSSONETIA L'Heritier 

BROUSSONETIA PAPYRI FERA Vent. Tabl. Regn. Veg. 3 (1794) 547. 
Frutex lintearius Rumph. Herb. Amb. 4: 114, t. 53. 

This reduction was made by Henschel, Vita Rumph. (1833) 
165, and is apparently the correct disposition of Frutex lin- 
tearius. The material on which the figure and the description 
were based was from Celebes, The figure is poor and presents 
only a branch with leaves; the flowers and the fruits are not 
described. 

TAXOTROPHIS Blume 

TAXOTROPHIS ILICIFOLIA Vidal Rev. PI. Vase. Filip. (1886) 249. 

Balanostreblus ilicifolia Kurz in Journ. As. Soc. Beng. 42 1 (1873) 
247, t. 19. 

Taxotrophis triapiculata Gamble Kew Bull. (1913) 188. 
Taxotrophis obtusa Elm. Leafl. Philip. Bot. 5 (1913) 1813. 
Ulet Rumph. Herb. Amb. 3: 62, t. 3J>. 

No previous reduction of Ulet has been suggested other than 
Hasskarl's suggestion that it might be an Antidesma. It is, 
however, unquestionably a form of Taxotrophis ilicifolia Vidal, 
as shown by the figure and the description. In regard to the 
synonymy cited above, Taxotrophis ilicifolia Vidal was published 



MORACEAE 



189 



independently of Balanostreblus ilicifolia Kurz. Kurz's species 
has been reported from Celebes by Koorders, and an examination 
of Koorders's specimens at Buitenzorg convinced me that they 
were the same as the Philippine Taxotrophis ilicifolia Vidal, 
which lead me to suspect that Balanostreblus ilicifolia Kurz and 
Taxotrophis ilicifolia Vidal were one and the same thing. I 
am indebted to Sir D. Prain, director of the Royal Gardens, 
Kew, England, for the following memorandum supplied to me 
under date of June 28, 1916: 

The material of Balanostreblus ilicifolius Kurz, and Taxotrophis ilici- 
folia Vidal has been compared in accordance with your request of May 
9th, with the result that your surmise as to their identity is very probably 
correct. There is a slight difference in the male catkins which may be of 
no importance. In the Philippine plant they are very short, while in 
Balanostreblus ilicifolius, even in a young state, they are much longer. 
Balanostreblus Kurz, will therefore have to be reduced to Taxotrophis 
Blume, the anthers of which appear to be inflexed. I might mention that 
Taxotrophis triapiculata Gamble (Kew Bull. 1913, 188) proves to be the 
same as Balanostreblus ilicifolius Kurz. 

I am now of the opinion that the recently described Taxotro- 
phis obtusa Elm. is also a form of T. ilicifolia Vidal. 

CUDRANIA Trecul 

CUDRANIA JAVANENSIS Tree, in Ann. Sci. Nat. Ill 8 (1847) 123. 
Trophis spinosa Blume Bijdr. (1825) 489, non Roxb. 
Batis spinosa Roxb. Hort. Beng. (1814) 71 (type!), non Fl. Ind. ed 

2, 3 (1832) 762. 
Cudranus rumphii Thw. Enum. PI. Ceyl. (1861) 262. 
Madura amboinensis Blume Mus. Bot. 2 (1849) 83. 
Cudranus amboinensis Miq. Fl. Ind. Bat. 1 2 (1859) 290. 
Cudranus spinosus O. Kuntze Rev. Gen. PI. 2 (1891) 623. 
Cudrania spinosa Hochr. in Bull. N. Y. Bot. Gard. 6 (1910) 489. 
Cudranus bimanus Rumph. Herb. Amb. 5: 22, t. 15, f. 2. 
Cudranus amboinicus Rumph. Herb. Amb. 5: 22, t. 15, /. 1. 
Cudranus amboinicus silvestris Rumph. Herb. Amb. 5: 25, t. 16. 

Not represented in our Amboina collections. The three 
"species" described and figured by Rumphius are apparently all 
referable to Cudrania javanensis Tree, which name I interpret 
as the oldest valid one for the species. Cudranus bimanus 
Rumph. is the type and whole basis of Batis spinosa Roxb. as 
published in the Hortus Bengalensis (1814) 71,* but I consider 
the name to be invalidated in Cudrania by C. spinosa (Blume) 
Hochr. The species later described by Roxburgh under this 
name is entirely different from Cudrania javanensis Tree. 



* See C. B. Robinson in Philip. Journ. Sci. 7 (1912) Bot. 415. 



190 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Trophis spinosa Roxb. as published by Willdenow, Sp. PL 4 
(1805) 734, to which Cudranus bimanus Rumph. was also re- 
duced, is Plecospermum spinosum Tree. Retzius, Obs. 5 (1789) 
30, reduced it to Trophis aspera Retz. with doubt, where it mani- 
festly does not belong. Loureiro, Fl. Cochinch. (1790) 548, 
discussed the form figured on t. 16 under Morella rubra Lour.— 
Myrica nagai Thunb., which led Poiret to cite it under Asca- 
rina rubra Poir., in Lam. Encycl. Suppl. 1 (1810) 475. The 
figures given by Rumphius have been cited under one or another 
of the various synonyms listed above. 

ARTOCARPUS Forster 

ARTOCARPUS INTEGRA (Thunb.) comb. nov. 

Radermachia Integra Thunb. in Vet. Akad. Handl. Stockh. (1776) 254. 
Artocarpus integrifolia Linn. f. Suppl. (1781) 412. 
Polyphema jaca Lour. Fl. Cochinch. (1790) 546. 
Artocarpus jaca Lam. Encycl. 3 (1789) 209. 

Soccus (Saccus) arboreus major Rumph. Herb. Amb. 1: 104, t. 30. 

The common jak fruit is well figured by Rumphius. Both 
t. 30 and the next species, t. 31, were reduced to Artocarpus 
integrifolia Linn. f. in the original description of that species, 
which was, however, essentially based on Radermachia integra 
Thunb. Following the rule of priority, I have here accepted 
Thunberg's specific name. 

ARTOCARPUS CHAMPEDEN (Lour.) Spreng. Syst. 3 (1826) 804. 
Polyphema champeden Lour. Fl. Cochinch. (1790) 547. 
Artocarpus polyphema Pers. Syn. 2 (1805) 531. 
Soccus (Saccus) arboreus minor Rumph. Herb. Amb. 1: 107, t. 31. 

The Rumphian species was cited by Loureiro in the original 
description of Polyphema champeden Lour., which is the basis 
of both Artocarpus polyphema Pers. and A. champeden Spreng. 
The figure has been cited by several authors under Artocarpus 
integrifolia Linn. f. or as a variety of that species. The species 
commonly called Artocarpus polyphema Pers., here, following 
priority, called Artocarpus champeden Spreng., is manifestly the 
one intended by Rumphius. 

ARTOCARPUS COMMUNIS Forst. Char. Gen. (1776) 101. 

Radermachia incisa Thunb. in Vet. Akad. Handl. Stokh. (1776) 254. 
Artocarpus incisa Linn. f. Suppl. (1781) 411. 
Soccus lanosus Rumph. Herb. Amb. 1 : 110, t. 32. 
Soccus granosus Rumph. Herb. Amb. 1: 112, t. 33. 

Soccus lanosus and Soccus granosus are respectively the seed- 
less and seeded forms of the breadfruit, corresponding to the 
forms described by Blanco as Artocarpus rima Blanco and A. 



MORACEAE 



191 



camansi Blanco. They are both referable to Artocarpus com- 
munis Forst. as that species is currently interpreted. Both, 
together with Soccus silvestris Rumph., were reduced to Arto- 
carpus incisa Linn. f. i • the original description of that species, 
which is typified by Radermachia incisa Thunb. 

ARTOCARPUS ELASTICA Reinw. ex Blume Bijdr. (1825) 481. 
Soccus silvestris Rumph. Herb. Amb. 1: 114, t. 34? 

Not represented in our Amboina collections. This has been 
reduced to Artocarpus communis Forst. (A. incisa Linn, f.) by 
several authors and may be a sylvan form of that species, or 
it may prove to be referable to Artocarpus elastica Reinw., where 
it was placed by Teysmann as quoted by Hasskarl, Neue Schlussel 
(1866) 16. 

ARTOCARPUS RETICULATA Miq. Ann. Mus. Bot. Lugd. Bat. 3 (1867) 
213. 

Novella cinerea Rumph. Herb. Amb. 2: 227? 
Amboina, Wae, Robinson PL Rumph. Amb. 173, November 26, 1913, 
along the seashore, locally known as mulewan. 

The specimen agrees quite closely with Miquel's description 
of Artocarpus reticulata, but its identity with Novella cinerea 
Rumph. is rather problematical. 

ARTOCARPUS FRET I SSI Teysm. & Binn. ex Hassk. in Abhandl. Nuturf. 
Gesellsch. Halle 9 (1866) 189 (Neue Schlussel 47) (type!). 
Metrosideros spuria I, mas Rumph. Herb. Amb. 3: 26. t. 13, f. A. 

A species of entirely doubtful status, to be interpreted from 
the description and figure given by Rumphius. As published, 
Artocarpus fretissi Teysm. & Binn. is typified wholly by the 
reference to Rumphius, as no description of the species was 
published by Teysmann and Binnendyck; the name does not 
appear in Index Kewensis. It is possible that the specimen 
intended by Teysmann and Binnendyck to represent the species 
is the one collected in Amboina by De Fretes and cited by Miquel, 
Ann. Mus. Bot. Lugd. Bat. 3 (1867) 213, under Artocarpus 
lakoocha Roxb. 

ARTOCARPUS sp. 

Metrosideros spuria II femina Rumph. Herb. Amb. 3: 27, t. 13, f. B. 

The description and figure are apparently those of an Arto- 
carpus, but a further determination of its status is impossible 
at this time from the material and data available. 

ARTOCARPUS sp. 

Soccus silvestris celebicus Rumph. Herb. Amb. 1: 115. 

The description is hardly sufficient to warrant an attempt 



192 



RUMPHIUS'S HERBARIUM AMBOINENSE 



at determining the status of this form. Henschel's suggestion 
that it might be Artocarpus lakoocha Roxb. is not tenable. 

ARTOCARPUS sp. 

Caju bandaa Rumph. Herb. Amb. 1: 109. 

As to the reference to Rheede this is Artocarpus hirsuta Lam., 
but the Javan plant discussed must be a different species, and 
its status is indeterminable at this time. 

ANTIARIS* Leschenault 

ANTIARIS TOXICARIA (Pers.) Lesch. in Ann. Mus. Paris 16 (1810) 478. 

Ipo toxicaria Pers. Syn. 2 (1807) 566. 

Arbor toxicaria mas Rumph. Herb. Amb. 2: 263, t. 87. 

This is the famous upas tree, and it is manifestly the form 
commonly known as Antiaris toxicaria Lesch. Arbor toxicaria 
Rumph. was reduced to Ipo toxocaria Pers. in the original 
description of the genus and species. 

The form described by Rumphius as Arbor toxicaria femina is 
probably referable here. Blume has placed it under Antiaris 
innoxia BL, which is a synonym of A. toxicaria Lesch. 

FICUS Linnaeus 

FICUS RACEM I FERA Roxb. Hort. Beng. (1814) 66 (type!). 

Ficus amboinensis Kostel. Allgem. Med.-Pharm. Fl. 2 (1833) 408 
(type!). 

Ficus nodosa Teysm. & Binn. in Nat. Tijdschr. Ned. Ind. 29 (1867) 
245. 

Caprificus amboinensis esculenta latifolia Rumph. Herb. Amb. 3: 145, 
t. 93. 

Amboina, Gelala and vicinity of the town of Amboina, Robinson PI. 
Rumph. Amb. 181, July and August, 1913, along streams at low altitudes, 
locally known as gondal. 

The Rumphian species was originally reduced by Linnaeus 
through error to Ficus benghalensis Linn., in Stickman Herb. 
Amb. (1754) 13, Amoen. Acad. 4 (1759) 124. Loureiro, Fl. 
Cochinch. (1790) 665, discusses it under Ficus auriculata Lour., 
which, according to his description, is an entirely different 
species. It is the type and whole basis of Ficus racemifera 
Roxb. as originally published in the Hortus Bengalensis (1814) 
66, by citation of the Rumphian figure, but is not the form later 
described by Roxburgh, Fl. Ind. ed. 2, 3 (1832) 560 (=Ficus 
variegata Blume), where the reference to Rumphius is repeated. 
It is also the type of Ficus amboinensis Kostel., which thus 
becomes a synonym of Ficus racemifera Roxb. The type of 



* Retained name, Vienna Code; Ipo Pers. (1807) is older. 



MORACEAE 



193 



Ficus nodosa Teysm. & Binn. was from Amboina. The figure 
given by Rumphius is decidedly poor, and from it alone the 
status of the species is indeterminable; in connection with 
Amboina material, however, it is clearly the species as here 
interpreted. 

FICUS MOSELEYANA King in Ann. Bot. Gard. Calcutta 1 (1888) 144, 
t. 181. 

Caprificus aspera tertia Rumph. Herb. Amb. 3: 151. 
Amboina, Mahija, Robinson PL Rumph. Amb. 181, August 7, 1913, 
margins of forests, altitude about 250 meters, locally known as koti and 
gohi. 

There is some doubt as to the correctness of this reduction of 
the Rumphian name, for the form described as Caprificus aspera 
III may be properly referable to Ficus wassa Roxb. 

FICUS WASSA Roxb. Fl. Ind. ed. 2, 3 (1832) 539. 

Caprificus aspera latifolia Rumph. Herb. Amb. 3: 150, t. 94.. 

Caprificus aspera angustifolia Rumph. Herb. Amb. 3: 151. 
Amboina, Soja, Elephant River, and town of Amboina, Robinson PL 
Rumph. Amb. 175, 176, 177, July, August, and September, 1913, from sea 
level to an altitude of 400 meters, locally known as gohi. 

Roemer and Schultes erroneously reduced the Rumphian 
species to Ficus symphiti folia Lam. ; Pritzel erroneously referred 
it to Ficus glomerata Roxb.; and Haskarl, Neue Schliissel (1866) 
60, thought that it might be Covella hispida Miq. The type of 
Ficus wassa Roxb. was a specimen cultivated in the botanic 
garden at Calcutta, originating in the Moluccas, and Roxburgh 
states in the original description: "Wassa of the Malayas, and 
probably Caprificus aspera Rumph. Amb. Ill t. 9J>." The speci- 
mens agree closely with the description of Rumphius and of 
Roxburgh, but the plate given by Wight, Ic. t. 666, presents a 
specimen with much more prominently toothed leaves than our 
Amboina material and, for that matter, than Roxburgh's 
description calls for. The receptacles are both axillary and 
solitary and on short tubercle-like racemes on the branches and 
trunk. Wassa is one of the native names cited by Rumphius 
for this species. 

FICUS SEPTICA Burm. f. Fl. Ind. (1768) 226. 

Ficus leucantatoma Poir. in Lam. Encycl. Suppl. 2 (1811) 654. 
Ficus septica Rumph. Herb. Amb. 3: 153, t. 96. 
Amboina, Elephant River, near the town of Amboina, and Paso, Robinson 
PL Rumph. Amb. 189, July and October, 1913, locally known as siripopa. 

The specimen agrees entirely with Rumphius's figure and 
description, and also with Ficus leucantatoma Poir. as currently 

144971 ia 



194 



RUMPHIUS'S HERBARIUM AMBOINENSE 



interpreted. Ficus septica Burm. f. has been quite overlooked 
by recent authors, but I believe that this name should be adopted 
in place of Poiret's. The Rumphian figure and description have 
been cited under Ficus septica Burm. f., by Loureiro, Lamarck, 
Vahl, Roemer and Schultes, Henschel, Kosteletzky, Walpers, 
Pritzel, and Miquel [Fl. Ind. Bat. I 2 (1858) 311], but the name 
is not included in the later writings of Miquel on Ficus [Ann. 
Mus. Bot. Lugd. Bat. 3 (1867) 260-300], nor by King in his 
monumental work on the species of Ficus of the Indo-Malayan 
region [Ann. Bot. Gard. Calcutta 1 (1888) 1-185, t. 1-232]. In 
the original description Burman f. first cited the Rumphian 
synonym, taking his specific name from Rumphius, followed by a 
citation of Handur-alu Rheede, Hort. Malabar. 3 : 77, t 59, the 
citation of the Javanese name siri bipar, and the statement 
"Habitat in India;" it seems to be quite evident that he had 
a Javan specimen. 

FICUS ALTISSI M A Blume Bijdr. (1825) 455. 

Varinga latifolia Rumph. Herb. Amb. 3: 127, t. Sk bis. 

This is not represented in our Amboina collections. It was 
originally reduced by Linnaeus to Ficus racemosa Linn., in Stick- 
man Herb. Amb. (1754) 13, Amoen. Acad. 4 (1759) 124, where 
it certainly does not belong, and later, ^Sp. PI. ed. 2 (1763) 1515, 
was placed as a variety of Ficus indica Linn., which it certainly 
does not represent, no matter whether the latter be interpreted 
from the occidental or the oriental references. Several authors 
.followed Linnaeus in citing Varinga latifolia Rumph. under Ficus 
indica Linn., and Roxburgh, Fl. Ind. ed. 2, 3 (1832) 539, states 
"the figure very bad" (i. e. for Ficus indica Linn.) . Vahl, Enum. 
2 (1805) 189, erroneously places it under Ficus cotoneaefolia 
Vahl. The figure is not good, and the data given in the descrip- 
tion indicate that it is greatly reduced, so that it somewhat 
resembles Ficus gelderi Miq. The description, however, applies 
very closely to Ficus altissima Blume and certainly represents 
this species or a very closely allied one. 

FICUS PUNCTATA Thunb. Ficus (1786) 9. 

Crusta arborum minor Rumph. Herb. Amb. 5: 84, t. 45. 
Amboina, Amahoesoe and Hoetoemoeri road, Robinson PI. Rumph. Amb. 
182, August and September, 1913, climbing on trees, altitude 4 to 100 
meters, locally known as tali oit. 

No previous reduction of this Rumphian species has been sug- 
gested, other than Hasskarl's reference of it to the genus Ficus. 
The figure, the description, and the Amboina specimen cited 



MORACEAE 



195 



above, all manifestly represent a form of the widely distributed 
Ficus punctata Thunb. The forms described in this chapter as 
Crusta arborum II alba, III odorata, and IV minima, with the possible 
exception of the last, are species of Ficus, apparently all different 
from Ficus punctata Thunb., but their more exact status is 
quite indeterminable from the data given by Rumphius. 

FICUS CONORA King in Ann. Bot. Gard. Calcutta 1 (1888) 103, t. 181. 
Caprificus viridis major Rumph. Herb. Amb. 3: 152, t. 85. 
Amboina, Way tommo and Hitoe lama, Robinson PL Rumph. Amb. 186, 
187, August and October, 1913, in forests, ascending to an altitude of 20 
meters, locally known as mussor. 

The figure is a good representation of Ficus conora King, which 
is found in the Philippines, Ternate, and New Guinea; and the 
specimens agree with the figure, with the description, and also 
in the native name as cited by Rumphius. Blume erroneously 
reduced it to Ficus vibes Reinw., an allied species with smaller 
leaves and very much smaller receptacles. Henschel erroneously 
places it under Ficus hispida Blume, while Miquel thought that 
it might be Covellia congesta Miq., perhaps from Roxburgh's dis- 
cussion of the Rumphian figure under Ficus congesta Roxb., Fl. 
Ind. ed. 2, 3 (1832) 560. 

FICUS RUM PHI I Blume Bijdr. (1825) 437. 

Ficus cordifolia Roxb. Fl. Ind. ed. 2, 3 (1832) 548, non Blume. 

Arbor conciliorum Rumph. Herb. Amb. 3: 142, t. 91, 92. 
Amboina, Hitoe lama, Robinson PI. Rumph. Amb. 180, November 5, 1913, 
along the seashore. 

Arbor conciliorum Rumph. was originally reduced by Linnaeus 
to the allied Ficus religiosa Linn., in Stickman Herb. Amb. 
(1754) 13, Amoen. Acad. 4 (1759) 124, Syst. ed. 10 (1759) 1315, 
Sp. PL ed. 2 (1763) 1514. Blume, however, placed it under 
Ficus rumphii Bl. in the original description of that species, 
and Roxburgh likewise cites it in the original description of 
Ficus cordifolia Roxb. It was erroneously placed by Roemer 
and Schultes under Ficus populnea Willd., an American species. 

FICUS BEN J AM I N A Linn. Mant. 1 (1767) 129. 

Varinga parvifolia Rumph. Herb. Amb. 3: 139, t. 90. 
Amboina, Paso, Robinson PI. Rumph. Amb. 179, October 31, 1913, along 
the seashore, locally known as waringin. 

Ficus benjamina Linn., as originally published, manifestly 
includes more than one species, but I have here followed the 
current interpretation of it. Varinga parvifolia Rumph. was 



196 



RUMPHIUS'S HERBARIUM AMBOINENSE 



quoted as a synonym in the original description of the species, 
but is not the type. 

Two forms are described by Rumphius in this chapter ; namely, 
Varinga parvifolia alta, which is the one figured and here inter- 
preted as Ficus benjamina, and Varinga parvifolia II humilis, smaller 
in size and with somewhat larger leaves and larger fruits than 
the former. Blume thought that this might be Ficus haemato- 
carpa Blume, while Hasskarl placed it with doubt under 
Urostigma neglectum Miq. Its exact status is indeterminable 
from the data and the material at present available for study. 

FICUS TREM ATOCARPA Miq. Ann. Mus. Bot. Lugd. Bat. 3 (1867) 224. 
Grossularia domestica Rumph. Herb. Amb. 3: 136, t. 87, 88. 
Amboina, Kati-kati, Robinson PL Rumph. Amb. 178, October 7, 1913, 
in light forests, altitude about 80 meters, locally known as waringin daun 
alus. I am disposed to refer to the same species Rel. Robins. 1680, 1681, 
1682, from Wae, Paso, and Batoe gadjah, all indicated as waringin. 

Grossularia domestica Rumph. was erroneously referred by 
Linnaeus to Ficus racemosa Linn., in Stickman Herb. Amb. 
(1754) 13, Amoen. Acad. 4 (1759) 124; the only other suggested 
reduction is Miquel's doubtful reference of it to Ficus altimeraloo 
Roxb. The probabilities are very great that it is here correctly 
referred to Ficus trematocarpa Miq., the type of which was from 
Amboina, although the status of Miquel's species is doubtful. 
King has reduced it with several other species, including the 
Philippine Ficus philippinensis Miq., to Ficus decaisneana Miq., 
which disposition of it is perhaps correct; but if King's concep- 
tion of the specific limits be correct, then it is probable that 
several other described Philippine forms will have to be reduced, 
such as Ficus inaequifolia Elm., F. confusa Elm., F. driveri Elm., 
F. setibracteata Elm., and F. magnifica Elm. 

FICUS ADENOSPERMA Miq. Ann. Mus. Bot. Lugd. Bat. 3 (1867) 233. 
Caprificus viridis minor Rumph. Herb. Amb. 3: 152. 
Amboina, Way tommo, and Roemah tiga, Robinson PL Rumph. Amb. 
18U, 185, August, 1915, along streams at low altitudes, locally known as 
kaju musor. 

The specimens agree fairly well with Rumphius's description 
and perfectly with that of Miquel. The type of Ficus adeno- 
sperma Miq. was from Amboina. 

FICUS AMPELOS Burm. f. Fl. Ind. (1768) 226 p. p. 

Folium politorium vulgare fruticosum Rumph. Herb. Amb. 4: 128. 
t. 63. 

This particular form is not represented in our Amboina collec- 



MORACEAE 



197 



tions, but the plant figured and described by Rumphius is 
apparently the same as the Javan form currently referred to 
Ficus ampelos Burm. f. The Rumphian description and figure, 
cited by Burman f. in the original description of Ficus ampelos 
Burm. f., should probably typify the species. It has been re- 
duced to Ficus politoria Lam., but Lamarck's species, based on 
specimens from Madagascar, is certainly distinct from the 
Malayan form. It has also been erroneously referred to Ficus 
parasitica Roth, and to F. exasperata Roxb. 

FICUS CORONATA Reinw. ex Blume Bijdr. (1825) 470. 
Ficus obscura Blume 1. c. 474. 

Folium politorium arborescens Rumph. Herb. Amb. 4: 128. 
Amboina, Kati-kati, Robinson PI. Rumph. Amb. 183, October 18, 1913, 
borders of clearings, altitude about 70 meters, locally known as daun plas. 

This form, with very scabrid, somewhat inequilateral leaves, 
certainly represents Folium politorium arborescens Rumph. 
I cannot distinguish it from Ficus coronata Reinw. (F. obscura 
Blume). The form described by Rumphius, 1. c, as Folium 
politorium flagellare is probably referable to one or the other of the 
above species with harsh leaves. 

FICUS FORSTEN 1 1 Miq. Ann. Mus. Bot. Lugd. Bat. 3 (1867) 214, 285. 
Varinga supa Rumph. Herb. Amb. 3: 135, t. 86? 

Nothing resembling this form occurs in our Amboina collec- 
tions. The figure very strongly resembles both Ficus pilosa 
Reinw. and F. forstenii Miq., and the form described by 
Rumphius is probably referable to one or the other of these 
species. Miquel thought that it might represent Urostigma 
pilosum Miq.=Ficus pilosa Reinw., but it seems to me that it more 
closely resembles Ficus forstenii Miq. Hamilton referred it with 
doubt to Ficus gonia Ham., and Henschel quite wrongly refers 
it to Ficus citrifolia Willd. The form described by Rumphius as 
Varinga pelal, in the same chapter, may be referable to Ficus 
forstenii Miq. or may represent a different species. 

FICUS sp. aff. F. CALOPHYLLA Blume. 

Varinga repens Rumph. Herb. Amb. 3: 134, t. 85. 

Nothing at all agreeing with Varinga repens Rumph. is 
presented by our Amboina collections. The figure and the 
description, however, apply closely to the Javan Ficus calophylla 
Blume and the Philippine Ficus pachyphylla Merr. and cer- 
tainly represent a species very closely allied to both and perhaps 
identical with one of them. Varinga repens Rumph. was erro- 
neously reduced by Linnaeus to Ficus pumila Linn., in Stickman 



198 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Herb. Amb. (1754) 13, Amoen. Acad. 4 (1759) 124, Syst. ed. 
10 (1759) 1315, Sp. PL ed. 2 (1763) 1515. Likewise it was 
erroneously reduced by Lamarck, with doubt, to Ficus pyrifolia 
Lam. ; by Vahl to Ficus rubra Vahl ; and by Blume to Ficus 
microcarpa Linn. f. var. litoralis Blume. Miquel thought that 
it might represent a species allied to Ficus manok Miq. 

FICUS RECURVA Blume Bijdr. (1825) 457. 

Rudens silvaticus parvifolius Rumph. Herb. Amb. 5 : 80, t. US, f. 2. 

From the description and figure I have little doubt in referring 
this to Ficus recurva Blume. A more comprehensive explora- 
tion of Amboina, however, may yield material that will modify 
this reduction. 

FORMS OF FICUS, DESCRIBED BY RUMPHIUS, OF DOUBTFUL STATUS 

Grossularia domestica longifolia Rumph. Herb. Amb. 3: 136. 
Grossularia domestica parvifolia Rumph. Herb. Amb. 3: 136. 

Both are perhaps forms of Ficus trematocarpa Miq. to which Grossularia 
domestica Rumph. pertains. 

Varinga funicularis Rumph. Herb. Amb. 3: 137. 

Varinga nounouck Rumph. Herb. Amb. 3: 137 (from Madagascar). 
Grossularia silvestris Rumph. Herb. Amb. 3: 138, t. 89. 

This form was erroneously placed by Lamarck under Ficus racemosa 
Linn., and by Henschel under Ficus tsjela Ham. Miquel thought that it 
might be near Ficus albinervia Miq. It is not represented in our Amboina 
collections. 

Arbor eusanda Rumph. Herb. Amb. 3: 141. 

Caprificus amboinensis esculenta angustifolia Rumph. Herb. Amb. 3: 146. 
Apparently a form allied to Ficus racemifera Roxb., as interpreted above, 

and to F. variegata Blume. 

Caprificus amboinensis esculenta silvestris Rumph. Herb. Amb. 3: 148. 
Caprificus amboinensis esculenta s. hahuol altera Rumph. Herb. Amb. 
3: 148. 

Caprificus s. sycomorus chartaria (amboinensis) Rumph. Herb. Amb. 
3: 149. 

Caprificus s. sycomorus chartaria (javanica) Rumph. Herb. Amb. 3: 149. 
Caju djurang (e Java) Rumph. Herb. Amb. 3: 151. 
Ficus septica silvestris Rumph. Herb. Amb. 3: 154. 
Ficus septica angustifolia Rumph. Herb. Amb. 3: 154. 

The last two are perhaps merely forms of Ficus septica Burm. f. (Ficus 
leucantatoma Poir.) as interpreted above, to which Ficus septica Rumph. 
pertains. 

Rudens silvaticus latifolius Rumph. Herb. Amb. 5: 80, t. J^S. f. 1. 

A scandent fig, apparently belonging in the group with Ficus recurva 
Blume. 

Rudens silvaticus rugosus Rumph. Herb. Amb. 5: 81. 

Manifestly *a species of Ficus and apparently belonging in the same group 
as the above. 



MORACEAE 



199 



Rudens silvaticus IV Rumph. Herb. Amb. 5: 80. 

Probably a species of Ficus, and perhaps the same plant as that very 
briefly described as Gummi susu Rumph., Herb. Amb. 5: 43. 

This list of indeterminable Ficus described by Rumphius, for the most 
part comprises those forms that are very inadequately described. Later 
some of them may be placed from the study of more specimens and data 
than are now available, but the list, from the standpoint of nomenclature 
and synonymy, is of no importance. 

CONOCEPHALUS Blume 

CONOCEPHALUS AM BOI N ENSIS (Zipp.) Warb. in Engl. Bot. Jahrb. 18 
(1894) 189. 

Poikilospermum amboinense Zipp. ex Miq. Ann. Mus. Bot. Lugd. Bat. 

1 (1863-64) 203. 
Funis muraenarum latifolia Rumph. Herb. Amb. 5: 68, t. 36. 

Amboina, Way tommo, Lateri, and Negri lama, Robinson PI. Rumph. 
Amb. 170, 172, August, 1913, climbing over trees at low and medium alti- 
tudes; probably referable here also is Robinson PI. Rumph. Amb. 171, from 
Hitoe lama, November 6, 1913, altitude about 75 meters. 

Linnaeus, in Stickman Herb. Amb. (1854) 19, Amoen. Acad. 4 
(1759) 129, erroneously referred plate 36, as Funis convolutus 
Rumph., to Melastoma octandrum Linn., doubtless by confusion 
with the species of Medinilla figured on the preceding plate. 
Hasskarl, Neue Schliissel (1866) 96, quotes Teysmann's opinion 
that the species figured represents a species of Conocephalus. 
I can see no reason for considering it other than Conocephalus 
amboinensis Warb., which was originally described by Zippel 
from Amboina material as a monotypic genus, Poikilospermum 
amboinense Zipp. 

CANNABIS Linnaeus 

CANNABIS SATIVA Linn. Sp. PI. (1753) 1027. 

Cannabis indica Lam. Encycl. 1 (1785) 695. 

Cannabis indica Rumph. Herb. Amb. 5: 208, t. 77, f. 1. 

Cannabis indica tertia Rumph. Herb. Amb. 5: 211, t. 77, f. 2. 

The plant figured and described by Rumphius as Cannabis 
indica is manifestly the common hemp. It was reduced by 
Linnaeus to Cannabis sativa Linn., in Stickman Herb. Amb. 
(1754) 21, Amoen. Acad. 4 (1759) 130, but by several of the 
early authors, following Lamarck, was cited under Cannabis 
indica Lam., a synonym of C. sativa Linn. Hasskarl, Neue 
Schliissel (1866) 112, makes Cannabis indica tertia Rumph. 
the type of a new variety, Cannabis sativa Linn. var. crispata 
Hassk., which, however, is apparently merely a variant of the 
common hemp. 



200 



RUMPHIUS'S HERBARIUM AMBOINENSE 



URTICACEAE 

LA PORTE A * Gaudichaud 

LAPORTEA AMPLISSIMA (Blume) Miq. Fl. Ind. Bat. 1 2 (1858) 232, 
Ann. Mus. Bot. Lugd. Bat. 4 (1869) 301. 

Urera amplissima Blume Mus. Bot. 2 (1859) /. 22 (without descrip- 
tion) . 

Folium urens latifolium Rumph. Herb. Amb. 3: 217, t. 

Amboina, Lateri, Soja, Negri lama, and Halong, Robinson PL Rumph. 
Amb. 310, 311, 312, 313, August and September, 1913, in forests and 
along rocky river banks, altitude 175 to 325 meters, locally known as 
polat, polot, and polat puti. 

Burman f., Fl. Ind. (1768) 205, referred this to Croton polot 
Burm. f., taking his specific name from Rumphius, but de- 
scribing the species from Javan material. Burman's species 
has been entirely overlooked by modern botanists and is the 
species commonly known as Claoxylon indicum Hassk., which 
should now be called Claoxylon polot (Burm. f.) {Croton polot 
Burm. f., Claoxylon indicum Hassk.). Henschel, followed by 
Pritzel, erroneously referred the Rumphian species to Jatropha 
moluccana Linn. =Aleur it es moluccana Willd. Hasskarl, Neue 
Schlussel (1866) 69, referred it to Laportea crenulata Gaudich... 
where it certainly does not belong, although he was correct as 
to the genus. In the original publication of Urera amplissima 
Blume no description is given; the name appears on the plate 
only. Miquel credits it to Java (Teysmann) and the Moluccas 
(Zippel), but in his second consideration of it he cites only 
Amboina material collected by Teysmann, Zippel, and Be Fretes, 
so that the original citation of Teysmann's specimen, as Javan, 
is probably an error. J. J. Smith f states : "De voor Java 
opgegeven L. amplissima Miq. is nog niet op Java aangetroffen, 
doch waarschijnlijk afkomstig van de Molukken. ,, 

LAPORTEA sp. 

Folium urens angustifolium Rumph. Herb. Amb. 3: 217. 
Folium urens rubrum Rumph. Herb. Amb. 3: 218. 

Both descriptions apparently apply to a single species of 
Laportea, which, from the data given by Rumphius, is perhaps 
distinct from Laportea amplissima Miq. Hasskarl, Neue 
Schlussel (1866) 69, thought that Folium urens rubrum Rumph. 

* Retained name, Vienna Code; Urticastrum Fabr. (1759) is older, 
f Koorders & Valeton Bijdr. Boomsoorten Java 12 (1910) 676. 



URTICACEAE 



201 



might be Laportea stimulans Miq. No. 310, cited above under 
Laportea amplissima Miq., was considered by Doctor Robinson 
certainly to represent Folium urens rubrum Rumph., and if 
this be correct, then this form described by Rumphius becomes 
a synonym of Laportea amplissima Miq. The leaf measure- 
ments given by Rumphius for all three forms are matched by 
some of the leaves on the specimens cited above. 

LAPORTEA DECUMAN A (Roxb.) Wedd. Monogr. Urt. (1856) 127. 

Urtica decumana Roxb. Hort. Beng. (1814) 67, nomen nudum, Fl. 

Ind. ed. 2, 3 (1832) 587. 
Urtica rumphii Kostel. Allgem. Med.-Pharm. Fl. 2 (1833) 400 (type!). 
Urtica decumana Rumph. Herb. Amb. 6: 47, t. 20, f. 1. 

Amboina, Halong and Hitoe lama, Robinson PL Rumph. Amb. 316, Sep- 
tember and October, 1913, along streams and in forests from near sea 
level to an altitude of 250 meters, locally known as daun gattal puti and 
daun gattal mera. 

This was originally reduced by Linnaeus, through error, to 
Urtica interrupta Linn., in Stickman Herb. Amb. (1754) 2G, 
Amoen. Acad. 4 (1759) 134. It has been cited by several 
authors under Fleurya interrupta Gaudich., but the figure does 
not pertain to this species, and the description only as to Urtica 
decumana III vulgaris; see below. Roxburgh cites the Rumphian 
description and figure in the original description of Urtica 
decumana Roxb., which was based on specimens introduced into 
the Calcutta Botanic Garden from the Moluccas. The Rumphian 
figure and description are the basis of Urtica rumphii Kostel. 
The figure is exceedingly poor, but it manifestly belongs with 
this species rather than with Fleurya interrupta Gaudich. I 
consider that the forms described by Rumphius as I alba and II 
rubra represent Laportea decumana (Roxb.) Wedd. as here 
interpreted. 

FLEURYA Gaudichaud 

FLEURYA INTERRUPTA (Linn.) Gaudich. Bot. Freyc. Voy. (1826) 497. 

Urtica interrupta Linn. Sp. PI. (1753) 985. 

Urtica decumana III vulgaris Rumph. Herb. Amb. 6: 48. 

Amboina, Robinson PI. Rumph. Amb. 319, July 19, 1913, in waste places, 
town of Amboina, locally known as daun gattal. 

There is very little doubt as to the correctness of this reduction 
of Urtica decumana vulgaris Rumph. The other plants de- 
scribed in the same chapter, Urtica decumana alba and rubra,, 
are apparently both Laportea decumana Wedd. 



202 



RUMPHIUS'S HERBARIUM AMBOINENSE 



PELL I ON I A Gaudichaud 

PELL I ON I A SINUATA (Blume) Boerl. Handl. Kenn. Fl. Nederl. Ind. 3 
(1900) 375. 
Procris sinuata Blume Bijdr. (1825) 511. 
Elatostema sinuatum Hassk. Cat. Hort. Bogor. (1844) 79. 
Macuerus mas Rumph. Herb. Amb. 6: 133, t. 58, f. 2. 
Amboina, Ayer putri, Robinson PI. Rumph. Amb. 320, July 28, 1913, on 
coral rocks at low altitudes. 

The only previously suggested identification of Macuerus mas 
is Hasskarl's doubtful reference of it to Elatostema macro- 
phyllum Brongn., Neue Schlussel (1866) 174. It is certainly 
not Brongniart's species, but is Pellionia sinuata Boerl., at 
least as that species is interpreted by Robinson, in Philip. Journ. 
Sci. 5 (1910) Bot. 497. 

BOEHMERIA Jacquin 

BOEHMERIA NIVEA (Linn.) Gaudich. Bot. Freyc. Voy. (1826) 499. 
Urtica nivea Linn. Sp. PI. (1753) 985. 
Rami urn majus Rumph. Herb. Amb. 5: 214, t. 79. f. 1. 

The plant figured and described is a form of ramie, probably 
typical Boehmeria nivea Gaudich., although possibly the variety 
tenacissima (Gaudich.) Miq. It was first reduced by Linnaeus 
to Urtica nivea Linn., in Stickman Herb. Amb. (1754) 21, 
Amoen. Acad. 4 (1759) 130, Sp. PI. ed. 2 (1763) 1398, and 
has been cited by various authors under Procris nivea Gaudich. 
and Boehmeria tenacissima Gaudich., synonyms of Boehmeria 
nivea Gaudich. Burman f., Fl. Ind. (1768) 197, erroneously 
referred it to Urtica aestuans Linn, in which he was followed by 
Lamarck, Persoon, and Henschel. 

POUZOLZIA Gaudichaud 

POUZOLZIA ZEYLANICA (Linn.) Benn. PI. Jav. Rar. (1838) 67. 

Parietaria zeylanica Linn. Sp. PI. (1753) 1052. 

Parietaria indica Linn. Mant. 1 (1767) 128. 

Pouzolzia indica Gaudich. Bot. Freyc. Voy. (1826) 503. 

Herba memoriae Rumph. Herb. Amb. 6: 29, t. 12, f. 2. 
Amboina, Batoe merah and vicinity of the town of Amboina, Robinson 
PI. Rumph. Amb. 324, 325, July, 1913, in sago swamps and along road- 
sides at low altitudes. 

Herba memoriae Rumph. is manifestly a form of Pouzolzia 
indica (Linn.) Benn. Burman f., Fl. Ind. (1768) 221, refers 
it to Parietaria indica Burm. f ., published independently of P. 
indica Linn. It has been cited as a synonym of Pouzolzia indica 
Gaudich. by several authors, for example, Presl, Bennett, Blume, 



URTICACEAE 



203 



and Miquel, but is here placed under what is manifestly its 
oldest valid specific name, Pouzolzia zeylanica (Linn.) Benn. 
Roxburgh, Fl. Ind. ed. 2, 3 (1832) 583, erroneously cites the 
Rumphian illustration under Urtica tuber -osa Roxb.= Pouzolzia 
tuber osa Wight. 

PIPTURUS Weddell 

PIPTURUS ARGENTEUS (Forst.) Wedd. in DC. Prodr. 16 1 (1869) 235". 

Urtica argentea Forst. Prodr. (1786) 65. 

Morus paniculata Roxb. Hort. Beng. (1814) 67 (type!). 

Perlarius I Rumph. Herb. Amb. 4: 120, t. 56. 
Amboina, Lateri, Batoe merah, and Amahoesoe, Robinson PL Rumph. 
Amb. 317, 318, July and August, 1913, in forests and thickets on limestone 
formations, altitude 20 to 150 meters, locally known as daun kes and 
daun kesi. 

Perlarius as figured and described by Rumphius is the whole 
basis of Morus paniculata Roxb. as originally published in the 
Hortus Bengalensis (1814) 67 by citation of the Rumphian 
figure; see C. B. Robinson in Philip. Journ. Sci. 7 (1912) Bot. 
414. Roxburgh's species, later described from material ori- 
ginating in the Moluccas, Fl. Ind. ed. 2, 3 (1832) 599, has been 
reduced to Pipturus velutinus Wedd.^P. incanus (Blume) 
Wedd., a species difficult to distinguish from P. argenteus 
Wedd. ; but wherever placed, it will in turn place the synonyms 
Pipturus paniculatus Miq. and Botry morus paniculata Miq. If, 
however, a critical revision of the genus should show that the 
Amboina material is Pipturus incanus Wedd. rather than P. 
argenteus Wedd., then Roxburgh's specific name will take pre- 
cedence over Pipturus incanus Wedd. 

Perlarius latifolius, described in this chapter, is Robinson- 
iodendron ambiguum Merr. (see p. 204). Perlarius parvifolius 
may be a variant of Pipturus argenteus Wedd. or may refer to 
some entirely different genus and species; its status is quite 
uncertain. 

PIPTURUS REPANDUS (Blume) Wedd. in Arch. Mus. Paris 9 (1857) 
448. 

Urtica repanda Blume Bijdr. (1825) 501. 
Aylaun nya femina Rumph. Herb. Amb. 5: 67. 
Amboina, Batoe gadjah and Batoe merah River, Robinson PI. Rumph. 
Amb. 321, 322, August, 1913, climbing over trees, ascending to an altitude 
of 250 meters. 

Aylaun nya femina Rumph. is described in the chapter with 
Funis muraenarum=Medinilla. The description is poor, but 
so far as it goes it applies fairly closely to the specimens here 
referred to Pipturus repandus Wedd. 



204 



RUMPHIUS'S HERBARIUM AMBOINENSE 



OREOCNIDE Miquel 

(Villebrunea Gaudichaud) 

OREOCNIDE RUBESCENS (Blume) Miq. Ann. Mus. Bot. Lugd. Bat. 4 

(1869) 306. 

Urtica rubescens Blume Bijdr. (1825) 506. 
Villebrunea rubescens Blume Mus. Bot. 2 (1856) 167. 
Lignum aquatile Rumph. Herb. Amb. 4: 135, t. 70. 

Amboina, Batoe merah River, Robinson PI. Rumph. Amb. 315, Septem- 
ber 24, 1913, on river banks, altitude about 40 meters. 

Hasskarl, Neue Schlussel (1866) 86, thought that Lignum 
aquatile Rumph. might be Oreocnide major Miq. or 0. silvatica 
Miq. ; according to J. J. Smith the former is a synonym of Ville- 
brunea rubescens B\ume=Oreocnide rubescens Miq., and the 
latter is a variety of it. The Amboina specimens, which agree 
closely with the figure, but not very well with the description, 
appear to be fairly typical of Oreocnide rubescens Miq. Rumph- 
ius's description is short and rather poor and may include more 
than this species. 

As to the genera Oreocnide Miquel and Villebrunea Gaudi- 
chaud, on a strict interpretation the former has priority; see 
C. B. Robinson in Philip. Journ. Sci. 6 (1911) Bot. 16. 

ROBI NSON IODEN DRON genus novum 

Genus Maoutiae affinis, differt perianthium 9 distincte evolu- 
tum, 5-denticulatum. 

ROBI NSON IODEN DRON AM BIGUUM (Wedd.) comb. nov. 

Maoutia ambigua Wedd. in Arch. Mus. Paris 8 (1855-56) 483; DC. 

Prodr. 16 1 (1869) 235 33 . 
Perlarius I latifolius Rumph. Herb. Amb. 4: 120. 

Amboina, Batoe merah, Robinson PI. Rumph. Amb. 21U, August 24, 
1913, on hillsides at low altitudes. 

The specimen cited above agrees perfectly with Rumphius's 
description, as well as with that of Maoutia ambigua Wedd., a 
species known only from Amboina and anomalous in the genus 
by its distinctly developed pistillate perianth. No previous 
reduction of Perlarius I latifolius Rumph. has been suggested, 
other than Hasskarl's opinion that it might be a variety of 
Perlarius I, that is, Pipturus argenteus Wedd. 

In view of the fact that the species is anomalous in Maoutia, 
where it was placed by Weddell, it seems best to establish 
another genus for it; in Maoutia the pistillate perianth is 
wanting. The generic name I have proposed is selected in 



PROTEACEAE-LORANTHACEAE 



205 



commemoration of Doctor Robinson's work in Amboina and 
of his critical work on the Philippine representatives of this 
difficult family.* 

PROTEACEAE 

HELICIA Loureiro 

H ELICI A SERRATA (R. Br.) Blume in Ann. Sci. Nat. II 1 (1834) 215. 

Rhopala serrata R. Br. in Trans. Linn. Soc. 10 (1811) 193. 

Arbor vespertilionum Rumph. Herb. Amb. 7: 17 (p. p., excl. t. 10!). 
Amboina, Paso, Robinson PL Rumph. Amb. 281, October 29, 1913, near 
the seashore. 

Arbor vespertilionum Rumph. was first reduced to Helicia 
serrata by Blume, in Ann. Sci. Nat. II 1 (1834) 215. He in- 
cluded also t. 10 with the statement "figura male expressa." 
The figure is manifestly no Helicia, refers to Arbor vesperti- 
lionum II described on page 17 following the description that does 
apply to Helicia, and is Schizomeria serrata Hochr. (see p. 244). 

LORANTHACEAE 
LOR AN THUS Linnaeus 

LORANTHUS RUM PHI I sp. nov. § Heteranthus. 

Viscum amboinicum album Rumph. Herb. Amb. 5: 60, t. 33. 
Amboina, Batoe gadjah and Soja road, Robinson PI. Rumph. Amb. 517 
(type), August 4, 1913; Koesoekoesoe sereh, Robinson PI. Rumph. Amb. 
516, August 23, 1913, parasitic, altitude 50 to 200 meters, locally known 
as bunga manumpang and daun manumpang. 

Frutex parasiticus glaber, ramis ramulisque tenuibus, tere- 
tibus; foliis oppositis, petiolatis, coriaceis, oblongis ad anguste 
oblongo-obovatis, obtusis, basi attenuatis, usque ad 8 cm longis. 
nervis utrinque 3 vel 4, obscuris vel subobsoletis, adscenden- 
tibus ; inflorescentiis axillaribus, solitariis, usque ad 2 cm longis ; 
floribus cylindraceis, teretibus, gracilis, circiter 2.4 cm longis. 
5- vel 6-meris, omnibus sessilibus, in triadibus vel diadibus 
breviter pedunculatis racemose dispositis. 

A parasitic, entirely glabrous shrub, the branches elongated, 
up to at least 70 cm in length, slender, terete, reddish-brown 
or dark in color when dry, the younger ones somewhat com- 
pressed at the nodes, smooth, the internodes 2 to 4 cm in length. 
Leaves opposite, coriaceous, olivaceous and rather dull when dry, 
oblong to narrowly oblong-obovate, 5 to 8 cm long, 1.5 to 3 cm 
wide, apex rounded to obtuse, base gradually narrowed, attenuate 

* Philippine Urticaceae. Philip. Journ. Sci. 5 (1910) Bot. 465-543; 6 
(1911) Bot. 1-33, t. 1-3; Philippine Urticaceae II, ibid. 299-314. 



206 



RUMPHIUS'S HERBARIUM AMBOINENSE 



or cuneate; lateral nerves 3 or 4 on each side of the midrib, 
ascending, slender, obscure, sometimes nearly obsolete; petioles 
about 8 mm long. Inflorescences axillary, solitary, about 2 cm 
long, the flowers slender, about 2.4 cm long, cylindric, orange 
below, yellowish above, 5- or mostly 6-merous, in racemosely 
arranged triads, the rachis slender, the lateral branchlets bearing 
the sessile flowers 2 to 4 mm long. Calyx cylindric-ovoid, 2 
to 3 mm long, truncate, the subtending bracts broadly ovate, 
rounded or obtuse, about 1.2 mm long, all the flowers sessile, 
usually in threes, sometimes in pairs at the apex of each short 
lateral branch. Petals free to the base, linear, about 1 mm wide, 
the reflexed portion above the insertion of the filament linear, 
6 to 7 mm long. Filaments 4 mm long, the anthers continuous, 
linear, about 1.2 mm in length. Fruit narrowly ovoid, when 
dry black and about 6 mm long. 

This species is possibly allied to Loranthus indicus Desr., but 
it differs in many characters, notably in its smaller leaves, longer 
petioles, and shorter, fewer-flowered inflorescences. The type 
of Loranthus indicus Desr. was a specimen collected by Sonnerat 
in "les Indies orientales," but de Candolle, Prodr. 4 (1830) 305, 
credits it to Timor. Miquel, Fl. Ind. Bat. I 1 (1856) 820, thought 
that Viscum amboinicum album Rumph. represented a species 
of Dendrophthoe (Loranthus) near D. indica (Desr.) Miq. or 
D. incarnata (Jack) Miq., but the Rumphian species is no 
Dendrophthoe, as indicated by its entirely free petals. 

Viscum amboinicum III Rumph. Herb. Amb. 5 : 62, very briefly 
described, is undoubtedly a species of Loranthus, but its status 
cannot be definitely determined from any data at present avail- 
able. Hasskarl, Neue Schliissel (1866) 95, thought it might 
be the same as Macrosolen macrophyllus Miq. (Loranthus ma- 
crophyllus Korth.), but this is entirely improbable, as that 
species is known only from Sumatra. 

ELYTRANTHE Blume 

ELYTRANTHE AM BOI N ENSIS sp. nov. 

Viscum amboinicum rubrum Rumph. Herb. Amb. 5: 61? 

Amboina, Hoetoemoeri road, Robinson PI. Rumph. Amb. 515 (type), 
September 30, 1913, on Barringtonia trees, altitude about 350 meters, 
locally known as manumpang. 

Frutex epiphyticus glaber, ramis crassis, ramulisque tere- 
tibus; foliis oppositis, coriaceis, nitidis, oblongo-ovatis, usque 
ad 18 cm longis, obtuse acuminatis, basi acutis ad subrotundatis, 
nervis utrique circiter 8, tenuibus, obscuris; inflorescentiis 



LORANTHACEAE 



207 



axillaribus, solitariis vel fasciculatis, brevibus, pedunculis 2- ad 
4-floris, circiter 5 mm longis; floribus 6-meris, circiter 2.5 cm 
longis, corollae tubo sursum prominente 6-angulato, bracteis late 
ovatis, circiter 1.5 mm longis, bracteolis paullo minoribus, con- 
natis, integris vel retusis. 

A stout, parasitic, glabrous shrub, the branches up to at 
least 60 cm in length, terete, brownish, somewhat lenticellate, 
the branchlets smooth, reddish-brown. Leaves opposite, thickly 
coriaceous, rather pale-greenish when dry, shining, oblong-ovate, 
or some of the smaller ones oblong-lanceolate, 14 to 18 cm long, 
4 to 7 cm wide, base acute to subrounded, apex shortly and 
broadly blunt-acuminate; lateral nerves about 8 on each side 
of the midrib, slender, anastomosing, more prominent on the 
upper than on the lower surface, the reticulations lax; petioles 
stout, 1 to 1.5 cm long. Inflorescence axillary, of solitary or 
fascicled, short peduncles up to 5 mm in length, each peduncle 
bearing at its apex from 2 to 4 short-pedicelled flowers, the 
peduncles 5 mm long or less, the pedicels about 2 mm in length. 
Flowers 6-merous, about 2.5 cm long, dark-red below, each 
subtended by one bract which is broadly ovate, acute, about 1.5 
mm long, the bracteoles entirely united or slightly refuse at 
the apex, nearly as large as the bract. Calyx cylindric, truncate, 
about 5 mm long. Corolla-tube about 6 mm long, inflated, prom- 
inently 6-angled or narrowly 6-winged in the upper one-third, 
the lobes 6, about 3 mm wide at the base, abruptly narrowed, 
about 14 mm long and 2 mm wide, thick, acute, spreading or 
reflexed. Filaments about 6 mm long, attached near the base 
of the lobes, the anthers continuous, linear, about 4 mm long. 

This species is sufficiently well characterized by its opposite, 
shining, rather large leaves, and especially by its axillary, soli- 
tary or fascicled, very short, 2- to 4-flowered peduncles. In Ely- 
tranthe it is distinguished by its bracteoles being either entirely 
connate into a single one nearly as large as the bract, or at 
most merely refuse at the apex. 

It is by no means certain that the plant here described repre- 
sents Viscum amboinense rubrum of Rumphius, as his short 
description does not well conform. Rumphius compares his 
plant with Viscum amboinicum album = Loranthus rumphii 
Merr., with which the present species has little in common. 
Viscum amboinense rubrum, moreover, was parasitic on small 
trees near the seashore, while Elytranthe amboinensis grows at 
an altitude of about 350 meters. Hasskarl, Neue Schliissel 
(1866) 95, suggests that Viscum amboinicum rubrum may 
be the same as Macrosolen evenius (Blume) Miq. = Loranthus 



208 



RUMPHIUS'S HERBARIUM AMBOINENSE 



evenius Blume, but in view of the characters of that species 
and its known distribution this proposed reduction is probably 
incorrect. 

SANTALACEAE 

SANTALUM Linnaeus 

SANTALUM ALBUM Linn. Sp. PI. (1753) 349. 
Sandalum Rumph. Herb. Amb. 2: 42, t. 11. 

The form described by Rumphius as Sandalum is probably 
the true sandal wood, Santalum album Linn., at least for most 
part. Sandalum was first reduced by Linnaeus, in Stickman 
Herb. Amb. (1754) 8, Amoen. Acad. 4 (1759) 120, Syst. ed. 10 
(1759) 1000, Sp. PL ed. 2 (1762) 497, to Santalum album Linn., 
which disposition of it has been very generally accepted by 
succeeding authors. Poiret, in Lamarck Encycl. 6 (1804) 502, 
suggested that it might be Sirium my rti folium Linn., which is 
generally cited as a synonym of Santalum album Linn. Prob- 
ably referable here is the form from Celebes indicated by 
Rumphius as Sandalum radicis Herb. Amb. 2 : 46, but there is less 
reason for considering Lignum papuanum III Rumph., 1. c. 58, as 
being identical with Santalum album Linn, as Henschel indi- 
cates; see Hasskarl, Neue Schliissel (1866) 28. 

EXOCARPUS LaBillardiere 

EXOCARPUS EPIPHYLLANTHUS (Linn.) comb. nov. 

Phyllanthus epiphyllanthus Linn. Amoen. Acad. 4 (1759) 136, Syst. 

ed. 10 (1759) 1264 (type!), non Sp. PL ed. 2 (1763) 1392. 
Xylophylla longifolia Linn. Mant. 2 (1771) 221 (type!). 
Phyllanthus cer amicus Pers. Syn. 2 (1807) 591 (type!). 
Exocarpus cer amicus R. Br. ex. Spreng. Gesch. 2 (1818) 77; Hensch. 

Vita Rumph. (1833) 201. 
Exocarpus phyllanthoides Endl. Prodr. Fl. Norfolk. (1833) 46? 
Exocarpos ceramica A. DC. Prodr. 14 (1857) 691 (type!). 
Xylophyllos ceramica Rumph. Herb. Amb. 7: 19, t. 12. 

The present application of Phyllanthus epiphyllanthus Linn, 
is entirely contrary to accepted usage, yet it is unquestionably 
the correct interpretation of the species. In the Amoen. Acad. 
4 (1759) 136, Linnaeus quotes the Rumphian illustration with 
doubt, but in the same year, Syst. ed. 10 (1759) 1264, he adds a 
description which is based wholly on Rumphius, as follows: 
u Phyllanthus epiphyllanthus. 2. P. f ol. lanceolatis serratis ; crenis 
floriferis. Rumph. amb. 7. t. 12." However, in the second edition 
of his Species Plantarum (1763) 1392, he discards the Rumphian 
synonym, adds various others, and describes a plant, a true 
Phyllanthus, from American material. Still later, Mant. 2 



OL AC ACE AE— ARISTOLOC H I ACE AE 



209 



(1771) 221, apparently recognizing his error, he proposed a 
new name for Phyllanthus epiphyllanthus of the second edition 
of the Species Plantarum, calling it Xylophylla latifolia Linn. 
A. de Candolle, Prodr. 14 (1857) 691, considers that Exocarpos 
ce? -arnica A. DC. is distinct from E. phyllanthoides Endl., but 
apparently saw no specimens of the former. Exocarpus 
rolfeanus (0. Kuntze) Merr., in Philip. Journ. Sci. 4 (1910) 
Bot. 352, is certainly very closely allied to E. epiphyllanthus 
(Linn.) Merr. and may have to be reduced to it. Of the names 
discussed above, Xylophylla longifolia Linn., Phyllanthus epi- 
phyllanthus Linn., as originally published, Phyllanthus ceramicus 
Pers., and Exocarpus ceramicus R. Br. and of A. de Candolle must 
all be interpreted from the Rumphian figure and description. 

OLACACEAE 

XI MEN I A Linnaeus 
XI MEN I A AMERICANA Linn. Sp. PL (1753) 1193. 

Zizyphus littorea Teysm. ex Hassk. in Abh. Naturf. Gesellsch. Halle 

9 (1866) 176 (type!). 
Vidara littorea Rumph. Herb. Amb. 2: 119, t. 37. 
Amboina, Paso, Robinson PL Rumph. Amb. 260, October 31, 1913, in 
thickets back of the beach. 

Vidara littorea Rumph. has not before been definitely placed; 
it is unquestionably the widely distributed strand plant, Ximenia 
americana Linn. It was reduced by Linnaeus, with doubt, to 
Rhamnus napeca Linn., in Amoen. Acad. 4 (1759) 121. Lour- 
eiro, Fl. Cochinch. (1790) 158, discusses it under Rhamnus 
soporifer hour. =Zizyphus soporifera Schultes. Hamilton, Mem. 
Wern. Soc. 6 (1832) 322, expresses the opinion that it is nearer 
to Elaeagnus than to Zizyphus, and Teysmann thought it a 
distinct species of Zizyphus which he called Zizyphus littorea, 
and which was published by Hasskarl, in Abh. Naturf. Gesellsch. 
Halle 9 (1866) 176 (Neue Schliissel 34). Of the names cited 
above, but a single one, Zizyphus littorea Teysm., is typified by 
the Rumphian figure, and this name is not listed in Index 
Kewensis. 

ARISTOLOCHIACEAE 

ARISTOLOCHIA Linnaeus 

ARISTOLOCHIA RUMPH II Kostel. Allg. Med.-Pharm. Fl. 2 (1833) 465 
(type!). 

Radix puluronica (s. Peponaster minor) Rumph. Herb. Amb. 5: 476, 
t. 177. 

This species is not represented in our Amboina collections. 

144971 14 



210 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Radix puluronica was originally reduced by Linnaeus to Aris- 
tolochia indica Linn., in Stickman Herb. Amb. (1754) 25, Amoen. 
Acad. 4 (1759) 133, Syst. ed. 10 (1759) 1249, in which he was 
followed by Loureiro, Henschel, and Pritzel, and, with doubt, 
by Duchartre, in DC. Prodr. 15 1 (1864) 479; this species, 
however, is definitely known only from India and Ceylon. I 
have here listed the Rumphian species under Aristolochia 
rumphii Kostel., of which it is the type, as it does not appear 
to be referable to any of the well-known Malayan species; such 
as Aristolochia tagala Cham., A. timoriensis Decne., A. gaudi- 
chaudii Duch., and A. zollingeri Miq. 

ARISTOLOCHIA sp. 

Peponaster major Rumph. Herb. Amb. 5: 474. 

Hasskarl, Neue Schliissel (1866) 149, thought that this might 
be the same as Aristolochia hastata Jack, but there is little 
reason to consider that this reduction is correct, as Jack's 
species is known only from Sumatra. It might be the very 
imperfectly described Aristolochia longifolia Roxb. (A. moluc- 
cana Duchartre), the type of which was from the Moluccas, 
or the very widely distributed Aristolochia tagala Cham. {A. 
roxburghiana Klotz.). Its status can be determined only by 
a critical study of all the Moluccan species when more abundant 
material is available. 

POLYGONACEAE 
RHEUM Linnaeus 

RHEUM RHABARBARUM Linn. Sp. PI. (1753) 372. 
Rhabarbarum sinense Rumph. Herb. Amb. 6: 148. 

The rather long discussion seems to apply to this Linnean 
species ; the plant itself is not described. Hasskarl, Neue Schliis- 
sel (1866) 177, referred it to Rheum undulatum Linn., which is 
a synonym of R. rhabarbarum Linn. 

RUM EX Linnaeus 

RUM EX PATIENTIA Linn. Sp. PI. (1753) 333. 

Lapathum hortense Rumph. Herb. Amb. 5: 277. 

Hasskarl, Neue Schliissel (1866) 120, suggests that this is 
Rumex patientia Linn., which is probably the correct disposition 
of it. The plant is not described by Rumphius, who merely 
states that it was the same as the European form called Acetosa 
hispanica, Pathic, or Patientia, that it was cultivated and used 
in cooking, and that it was known to the Malays as say or assam. 



CHENOPODIACEAE— AMARANTHACEAE 



211 



CHENOPODIACEAE 

CHENOPODIUM Linnaeus 

CHENOPODIUM QUINOA Willd. Sp. PI. 1 (1799) 1301. 
Blitum peruvianum Rumph. Herb. Amb. 5: 232. 

This South American species is briefly discussed. The reduc- 
tion, made by Hasskarl, to Chenopodium quinoa Willd., is prob- 
ably correct. Rumphius quotes the common name quinua for 
the species he discussed. 

SALICORNIA Tournefort 

SALICORNIA HERBACEA Linn. Sp. PL ed. 2 (1762) 5. 

Crithmus indicus III Kaly articulatum Rumph. Herb. Amb. 6: 166. 

This reduction follows Hasskarl, which is unquestionably the 
correct disposition of the European plant that Rumphius briefly 
discussed. 

AMARANTHACEAE 

DEERINGIA R. Brown 

DEERINGIA AM ARANTHOI DES (Lam.) comb. nov. 

Achyranthes amaranthoides Lam. Encycl. 1 (1785) 548. 
Celosia baccata Retz. Obs. 5 (1789) 23. 
Deeringia celosioides R. Br. Prodr. (1810) 413. 
Deeringia baccata Moq. in DC. Prodr. 13 2 (1849) 236. 
Blitum frutescens Rumph. Herb. Amb. 5: 235, t 83, f. 2. 

This widely distributed and well-known species is not repre- 
sented in our Amboina collections, but Rumphius's excellent 
figure is unmistakably the form commonly known as Deeringia 
celosioides R. Br. and as D. baccata Moq., but for which Achy- 
ranthes amaranthoides Lam. supplies an older name. Linnaeus, 
Sp. PI. ed. 2 (1762) 295, reduced Blitum frutescens Rumph. to 
Achyranthes muricata IArm.=Digera muricata (Linn.) Mart. 
The first and only citation in the original place of publication 
is the one to Rumphius, and this might by some authors be 
interpreted as the type of the species. However, the Linnean 
species is manifestly based primarily on an actual specimen and 
is hence not to be interpreted by the Rumphian reference. 
Lamarck, realizing that Linnaeus had confused two distinct 
species under Achyranthes muricata, proposed the name Achy- 
ranthes amaranthoides for what is now known as Deeringia 
baccata Moq., basing his description on specimens collected by 
Sonnerat, with the reduction of Blitum frutescens Rumph. The 
description applies unmistakably to Deeringia, not to Digera, 
although Lamarck's species has long been referred to Digera 



212 



RUMPHIUS'S HERBARIUM AMBOINENSE 



arvensis Forsk.=D. muricata (Linn.) Mart. Other names in- 
volved in the reduction of Blitum frutescens Rumph. are Clado- 
stachys arborescens Don, Cladostachys muricata Moq., and C. 
frutescens Don ; the first two are synonyms of Digera muricata 
(Linn.) Mart., the last is a synonym of Deeringia amaranthoides 
(Lam.) Merr. 

CELOSIA Linnaeus 

CELOSIA ARGENTEA Linn. Sp. PI. (1753) 205. 

Amarantus caudatus Rumph. Herb. Amb. 5: 237. 
Amboina, Koesoekoesoe sereh and town of Amboina, Robinson PL Rumph. 
Amb. 133, August, 1913, in waste places at low altitudes, locally known as 
bay am blanda. 

This reduction was first suggested by Hasskarl, Neue Schliissel 
(1866) 115, and I consider it to be the correct disposition of 
Amarantus caudatus Rumph. 

CELOSIA CRISTATA Linn. Sp. PL (1753) 205. 

Amarantus japonicus Rumph. Herb. Amb. 5: 236, t. 84. 
Amarantus vulgaris Rumph. Herb. Amb. 5: 236. 

This commonly cultivated species is not represented in our 
Amboina collections. Amarantus japonicus Rumph. was orig- 
inally reduced by Linnaeus to Celosia cristata Linn., in Stick- 
man Herb. Amb. (1754) 21, Amoen. Acad. 4 (1759) 130, Syst. 
ed. 10 (1759) 493, in which he was followed by numerous 
authors, and which is certainly the correct disposition of it. 
In the second edition of the Species Plantarum, however, (1762) 
297, he referred it to Celosia castrensis Linn., a synonym of 
Celosia cristata Linn. Hasskarl, Neue Schliissel (1866) 114, 
refers Amarantus japonicus Rumph. to Celosia cristata Linn, 
var. splendens Moq., and Amarantus vulgaris to Celosia cristata 
Linn. var. exaltata Hassk. 

AMARANTH US Linnaeus 

AMARANTHUS VIRIDIS Linn. Sp. PI. ed. 2 (1763) 1405. 

Blitum indicum domesticum Rumph. Herb. Amb. 5: 231, t. 82, f. 1. 

Blitum indicum I album Rumph. Herb. Amb. 5: 231. 

Blitum indicum II maculosum Rumph. Herb. Amb. 5: 231. 

Blitum indicum II maculosum amboinicum Rumph. Herb. Amb. 5: 231. 
Amboina, Robinson PI. Rumph. Amb. 139, August 20, 1913, in waste 
places, locally known as bayang. Also Rel. Robins. 2513, 2514 from Bali, 
July 7, 1913. 

Blitum indicum Rumph., including both forms figured by 
Rumphius on plate 82, was reduced by Linnaeus to Amaranthus 
tristis Linn., in Stickman Herb. Amb. (1754) 21, Amoen. Acad. 



AMARANTHACEAE 



213 



4 (1759) 130, Syst. ed. 10 (1759) 1268, Sp. PL ed. 2 (1763) 
1404, in which he was followed by various authors — Lamarck, 
Loureiro, Willdenow, Burman f., and others. The form de- 
scribed and figured appears to me to be Amaranthus viridis 
Linn. Blitum indicum I album Rumph. has also been reduced 
to Amaranthus polygamus Linn, by numerous authors, to Ama- 
ranthus oleraceus Linn., and to Euxolus polygamus Moq. 

AMARANTHUS GANGETICUS Linn. Syst. ed. 10 (1759) 1268. 

Blitum indicum IV terrestre Rumph. Herb. Amb. 5: 232, t. 82, f. 2? 
Amboina, Way tommo, Robinson PL Rumph. Amb. 136, August 16, 1913, 
locally known as bay a. 

The specimen cited appears to represent a dwarfed form of 
Amaranthus gangeticus Linn, and agrees fairly well with the 
figure cited. Blitum indicum IV terrestre Rumph. has been very 
generally reduced to Amaranthus tristis Linn. 

AMARANTHUS SPINOSUS Linn. Sp. PI. (1753) 991. 

Blitum spinosum Rumph. Herb. Amb. 5: 234, t. 83, f. 1. 
Amboina, Robinson PL Rumph. Amb. 137, 138, August, September, 1913, 
along road sides at low altitudes. 

The form cited aboVe certainly represents Blitum spinosum 
Rumph., but is not typical Amaranthus spinosus Linn., differing 
from it in its smaller flowers and in its few, scattered, short 
spines. The reduction was first made by Linnaeus, in Stickman 
Herb. Amb. (1754) 21, Amoen. Acad. 4 (1759) 130, Syst. ed. 
10 (1759) 1269, Sp. PL ed. 2 (1763) 1407, which disposition of 
it has been accepted by all authors. The figure is poor and is 
very greatly reduced in size. 

The form from Macassar, Celebes, very briefly described as 
Blitum spinosum e Macassar Rumph., Herb. Amb. 5: 234, is prob- 
ably the typical, robust form of Amaranthus spinosus Linn. 

AMARANTHUS TRICOLOR Linn. Sp. PI. (1753) 989. 
Amarantus versicolor Rumph. Herb. Amb. 5: 237. 
Blitum indicum III rubrum Rumph. Herb. Amb. 5: 232? 

The description of Amarantus versicolor applies unmistakably 
to the rather robust form with variegated and variously colored 
leaves found in cultivation in many typical countries. The 
species has been reduced by various authors to Amaranthus 
gangeticus Linn, and to A. melancholicus Linn., but Amaranthus 
tricolor Linn, has priority. The reduction of the form that 
Rumphius described was first made by Henschel. The form 
foliis obscure rubentibus described in this chapter by Rumphius 
probably also belongs here. 



214 



RUMPHIUS'S HERBARIUM AMBOINENSE 



CYATHULA Loureiro 

CYATHULA PROSTRATA (Linn.) Blume Bijdr. (1825) 549. i 
Achyranthes prostrata Linn. Sp. PI. ed. 2 (1762) 296. 
Cyathula geniculates Lour. Fl. Cochinch. (1790) 102. 
Auris canina I femina Rumph. Herb. Amb. 6: 26, t. 11. 
Amboina, Robinson PI. Rumph. Amb. 135, near the town of Amboina 

in a sago swamp at low altitude. 

Auris canina I femina Rumph. was originally reduced by- 
Linnaeus to Achyranthes lappacea Linn., in Stickman Herb. 
Amb. (1754) 26, Amoen. Acad. 4 (1759) 134, Syst. ed. 10 
(1759) 942, but recognizing his error, he later, Sp. PL ed. 2 
(1762) 296, cites it in the original description of Achyranthes 
prostrata lAxm.=Cyathula prostrata Blume. This is certainly 
the correct disposition of it. Other names involved in the 
reduction are Cyathula geniculata Lour., Desmochaete prostrata 
R. & S., and Pupalia prostrata Mart., all synonyms of Cyathula 
prostrata (Linn.) Blume. 

AERVA * Forskal 

AERVA SANGU INOLENTA (Linn.) Blume Bijdr. (1825) 547. 
Achyranthes sanguinolenta Linn. Sp. PI. ed. 2 (1762) 294. 
Illecebrum sanguinolentum Linn. Mant. 2 (1771) 344. 
Verbena rubra Rumph. Herb. Amb. 7: 60, t. 27, f. 2. 

This species is not represented in our Amboina collections. 
Verbena rubra Rumph. is cited by Linnaeus in the original 
description of Achyranthes sanguinolenta Linn., but the species 
was manifestly based on an actual specimen, because of the 
description added ; Verbena rubra Rumph., therefore, cannot be 
interpreted as the type of the species. This reduction of Ver- 
bena rubra Rumph. is certainly the correct disposition of it, and 
under one or the other of the above synonyms has been accepted 
by all authors. 

ACHYRANTHES Linnaeus 

ACHYRANTHES ASPERA Linn. Sp. PI. (1753) 204. 

Auris canina II mas Rumph. Herb. Amb. 6: 27, t. 12, f. 1. 

This common and well-known weed is not represented in our 
Amboina collections. Auris canina II mas was originally 
reduced by Linnaeus to Achyranthes aspera Linn., in Stickman 
Herb. Amb. (1754) 26, Amoen. Acad. 4 (1759) 134, Mant. 2 
(1771) 344, which is undoubtedly the correct disposition of it. 
Other authors, however, have referred it as follows: Lamarck 



* Retained name, Vienna Code; Comacum Adans. (1763) is older. 



NYCTAGINACEAE 



215 



to Achyranthes fruticosa Lam. ; Hasskarl to Achyranthes biden- 
tata Blume var. elongata Hassk. and to Achyranthes javanica 
Moq. 

ALTERNANTHERA Forskal 

ALTERNANTHERA SESSILIS (Linn.) R. Br. ex R. & S. Syst. 5 (1819) 
554. 

Gomphrena sessilis Linn. Sp. PL (1753) 225. 
Illecebrum sessile Linn. Sp. PL ed. 2 (1762) 300. 

Olus squillarum Rumph. Herb. Amb. 6: 37, t. 15, f. 1 (incl. / majus 
et 77 minus) . 

Amboina, Robinson PL Rumph. Amb. 134, near the town of Amboina, 
July and August, 1913, in ditches and in sago swamps at low altitudes. 

Linnaeus originally reduced Olus squillarum Rumph. to Gom- 
phrena sessilis Linn., in Stickman Herb. Amb. (1754) 24, 
Amoen. Acad. 4 (1759) 134, Syst. ed. 10 (1759) 950, which, 
as Alternanthera sessilis (Linn.) R. Br., is surely the correct 
disposition of it. Hasskarl, Neue Schliissel (1866) 158, thought 
that the description agreed better with Alternanthera nodi flora 
R. Br. than with A. sessilis R. Br. and that Olus squillarum II 
minus Rumph. was referable to Alternanthera nodiflora R. Br. 
var. lineari folia Moq. Alternanthera sessilis R. Br. is not 
published in the Prodromus (1810) 417, as currently indicated 
in botanical literature. 

GOMPHRENA Linnaeus 

GOMPHRENA GLOBOSA Linn. Sp. PL (1753) 224. 

Flos globosus Rumph. Herb. Amb. 5: 289, t. 100, f. 2. 

Flos globosus albus Rumph. Herb. Amb. 5: 290. 
■ Amboina, Negri lama, Robinson PI. Rumph. Amb. 132, September 8, 
1913, in fields at low altitudes, locally known as knop. 

This reduction of Flos globosus Rumph. was originally made 
by Linnaeus, in Stickman Herb. Amb. (1754) 22, Amoen. Acad. 
4 (1759) 131, is certainly the correct disposition of it, and has 
been universally accepted by authors who have had occasion 
to cite Rumphius's illustration and description. 

NYCTAGINACEAE 

MIRABILIS Rivinus 

M I RABI LIS JALAPA Linn. Sp. PL (1753) 177. 
Mirabilis Rumph. Herb. Amb. 5: 253, t. 89. 

This commonly cultivated plant is not represented in our 
Amboina collections. The figure given by Rumphius is an ex- 
cellent representation of this well-known species. The reduc- 



216 



RUMPHIUS'S HERBARIUM AMBOINENSE 



tion was first made by Linnaeus, in Stickman Herb. Amb. (1754) 
21, Amoen. Acad. 4 (1759) 130, Sp. PL ed. 2 (1762) 252. Seven 
color forms are included by Rumphius ; Hasskarl, Neue Schlussel 
(1866) 117, refers one of these to Mirabilis dichotoma Linn, 
and the others to several varieties of Mirabilis jalapa Linn. 

PI SON I A Plumier 

PI SON I A ALBA Spanoghe in Linnaea 15 (1841) 342. 
Olus album Rumph. Herb. Amb. 1: 191, t. 78. 

This commonly cultivated tree is not represented in our Am- 
boina collections. The reduction was apparently first suggested 
by Spanoghe in the original publication of Pisonia alba (not seen 
by me) ; repeated by Choisy, in DC. Prodr. 13 2 (1849) 446, as 
a doubtful synonym ; by Hasskarl referred with doubt to Pisonia 
morindae folia R. Br.; and in Retzia, 1 (1855) 6, as apparently 
representing Pisonia sylvestris Teysm. & Binn. The plant de- 
scribed by Rumphius is unmistakably the common and widely 
cultivated Malayan form with very pale-green or sometimes 
yellowish-white leaves and is typical Pisonia alba Spanoghe. 
Whether or not Pisonia alba Spanoghe is specifically distinct 
from P. sylvestris Teysm. & Binn. is uncertain, but the proba- 
bilities are that it is a derived form of P. sylvestris or of a 
closely allied species, which, through long cultivation, rarely 
produces flowers or fruits. In the Philippines it is generally 
known as coles moluco, certainly indicating that it was intro- 
duced into the Archipelago from the Moluccas. 

PISONIA GRAN DIS R. Br. Prodr. (1810) 422. 

Olus album insulare Rumph. Herb. Amb. 1: 190, t. 79, f. 1 (excl. 
fig. A, et descr. fruct.). 

No Pisonia, other than the very characteristic P. cauliflora 
Scheff., is represented in our Amboina collections, but the 
description given by Rumphius, excluding that of the fruit, is 
unmistakably applicable to the very large Pisonia that is fre- 
quently gregarious on small uninhabited islands in the Malayan 
and Polynesian regions, especially those frequented by birds, 
which has been described as Pisonia grandis R. Br. Hasskarl, 
Neue Schlussel (1866) 24, suggests that it may be Pisonia 
sylvestris Teysm. & Binn. which Heimerl considers to be a syn- 
onym of Pisonia grandis R. Br. 

PISONIA ACULEATA Linn. Sp. PI. (1753) 1026. 

Limonellus funicularis montanus Rumph. Herb. Amb. 5: 25. 

This reduction follows Hasskarl's suggestion, Neue Schlussel 
(1866) 91, the only objection being that the leaves, as described 



AIZOACEAE— PORTULACACEAE 



217 



by Rumphius, are decidedly larger than is the case with Pisonia 
aculeata. Blume, Bijdr. (1826) 735, places it under his Pisonia 
limonella and takes the specific name from Rumphius, but the 
actual type was a Javan plant; this is considered by all recent 
authors to be a synonym of Pisonia aculeata Linn. Choisy, in 
DC. Prodr. 13 2 (1849) 446, cites the Rumphian name under 
Pisonia limonellus Blume, but t. 16, also cited by him, does not 
belong with Limonellus funicular is montanus, but with Cudra- 
nus amboinensis sylvestris and is a Cudrania. 

AIZOACEAE 

SESUVIUM Linnaeus 

SESUVIUM PORTULACASTRUM Linn. Syst. ed. 10 (1759) 1058, Amoen. 
Acad. 4 (1759) 136. 
Portulaca portulacastrum Linn. Sp. PI. (1753) 446. 
Crithmus indicus I ruber Rumph. Herb. Amb. 6: 165, t. 72, f. 1. 
Crithmus indicus II albus Rumph. Herb. Amb. 6: 165. 

This widely distributed strand plant is not represented in 
our Amboina collections. The reduction of Crithmus indicus 
was first made by Linnaeus, in Stickman Herb. Amb. (1754) 
28, as Portulaca portulacastrum Linn., which as Sesuvium por- 
tulacastrum is manifestly the correct disposition of it. Many 
authors, however, have quoted it under Sesuvium repens Rottl., 
a synonym of S. portulacastrum Linn. 

The form mentioned as Crithmus indicus III kaly articulation 
Rumph. Herb. Amb. 6: 166, is manifestly the European Salicor- 
nia herbacea Linn., as placed by Hasskarl, Neue Schliissel 
(1866) 179. The form described as Crithmus indicus IV portulaca 
arenosa in the same chapter is apparently Portulaca quadrifida 
Linn., as placed by Hasskarl, 1. c. 

PORTULACACEAE 

PORTULACA Linnaeus 

PORTULACA OLERACEA Linn. Sp. PL (1753) 445. 

Portulaca indica Rumph. Herb. Amb. 5: 268 (incl. / major sativa, 
II rubra). 

Amboina, near the town of Amboina, Rel. Robins. 223, September 3, 
1913, locally known as rumput gelang mera. 

The identification of the Rumphian species was made by 
Hasskarl, Neue Schliissel (1866) 119, and is certainly the correct 
disposition of it. It is a widely distributed weed in all warm 
countries. 



218 



RUMPHIUS'S HERBARIUM AMBOINENSE 



PORTULACA QUADRIFIDA Linn. Mant. 1 (1767) 73. 

Portulaca indica Rumph. Herb. Amb. 5: 268 (incl. /// minima, IV 
litorea) . 

Hasskarl, Neue Schlussel (1866) 119, has reduced both of 
the above-mentioned forms, Portulaca indica III and IV, to 
P. quadrifida Linn. A single specimen of the latter occurs in 
our Amboina collection, PL Rumph. Amb. 222, from coral rocks 
at Silali, September 22, 1913. This may be the correct disposi- 
tion of the Rumphian plants, or they may have been merely 
forms of the commoner and variable Portulaca oleracea Linn. 

BASELLACEAE 

BASELLA Linnaeus 

BASELLA RUBRA Linn. Sp. PI. (1753) 272. 
Basella alba Linn. 1. c. 

Gandola I alba Rumph. Herb. Amb. 5: 417. 

Gandola II rubra Rumph. Herb. Amb. 5: 417, t. 154, f. 2. 

The form figured by Rumphius was originally reduced by 
Linnaeus to Basella rubra Linn., in Stickman Herb. Amb. (1754) 
24, Amoen. Acad. 4 (1759) 132, which has been accepted by 
all authors except Loureiro, Fl. Cochinch. (1790) 183, who 
called it Basella nigra Lour. ; the latter is manifestly a synonym 
of B. rubra Linn. The form described by Rumphius as Gandola 
alba was placed by Linnaeus, Sp. PL ed. 2 (1752) 390, under 
Basella alba Linn., which has very generally been followed by 
succeeding authors ; it is manifestly the form described by Lin- 
naeus as Basella alba Linn., but which is now reduced to B. 
rubra Linn., it being manifestly only a variant of that species. 

NYMPHAEACEAE 

NELUMBIUM Jussieu 

NELUMBIUM NELUMBO (Linn.) Druce Bot. Exch. Club (1914) 421. 
Nymphaea nelumbo Linn. Sp. PI. (1753) 511. 
Nelumbium speciosum Willd. Sp. PI. 2 (1799) 1258. 
Nelumbo javanica Poir. in Lam. Encycl. 4 (1798) 454. 
Nymphaea indica major Rumph. Herb. Amb. 6: 168, t. 73. 

Rumphius gives a good figure of the common lotus, this being 
first reduced by Linnaeus to Nymphaea nelumbo Linn., in Stick- 
man Herb. Amb. (1754) 28, Amoen. Acad. 4 (1759) 136, Syst. 
ed. 10 (1759) 1074, Sp. PL ed. 2 (1762) 730. It has been 
also cited under Nelumbium speciosum Willd., the commonly 
accepted name for the species, and Nelumbo javanica Poir., 
a synonym of Nelumbium nelumbo (Linn.) Druce. The several 



MENISPERMACEAE 



219 



color forms mentioned on page 169 probably do not pertain to 
the lotus, but to the genus Nymphaea. 

NYMPHAEA Linnaeus 
NYMPHAEA PUBESCENS Willd. Sp. PL 2 (1799) 1154. 
Castalia pubescens Blume Bijdr. (1825) 48. 

Nymphaea indica minor I vulgaris Rumph. Herb. Amb. 6: 172 p. p. 
Nymphaea indica minor III buronica Rumph. Herb. Amb. 6: 173. 

There is no representative of this genus in our Amboina 
collections. The color forms of Nymphaea indica major Rumph. 
are apparently not Nelumbium nelumbo (Linn.) Druce, but are 
referable here or in part to Nymphaea stellata Willd. Nymphaea 
indica I vulgaris has been referred by various authors to Nym- 
phaea lotus Linn., N. pubescens Willd., N. rubra Roxb., and N. 
stellata Linn. The description apparently includes both Nym- 
phaea pubescens Willd. and N. stellata Willd. Nymphaea indica 
II ceramica, the form figured, is Limnanthemum indicum Griseb. 
In regard to Nymphaea lotus Linn., the type is manifestly the 
African form, the original specimen still existing in the Linnean 
herbarium * where it was examined by Conard.f Fl. Zeyl. 19 U 
was wrongly placed here by Linnaeus as a synonym, and is 
Nymphaea pubescens Willd. 

NYMPHAEA STELLATA Willd. Sp. PI. 2 (1799) 1153. 
Castalia stellata Blume Bijdr. (1825) 49. 

Nymphaea indica minor I vulgaris Rumph. Herb. Amb. 6: 172, p. p. 

The color forms mentioned by Rumphius, for the most part, 
are apparently referable to Nymphaea stellata Willd. rather than 
to Nymphaea lotus Linn. As to the propriety of the use of 
Nymphaea instead of Castalia as the generic name for the water 
lilies, see Conard in Rhodora 18 (1916) 161. 

MENISPERMACEAE 

PERICAMPYLUS Miers 
PERICAMPYLUS GLAUCUS (Lam.) comb. nov. 

Menispermum glaucum Lam. Encycl. 4 (1797) 100 (type!). 
Cocculus glaucus DC. Syst. 1 (1818) 521 (type!). 
Cocculus incanus Colebr. in Trans. Linn. Soc. 13 (1822) 57, t. 17. 
Pericampylus incanus Miers in Ann. Nat. Hist. II 7 (1851) 40; Diels 

in Engl. Pflanzenreich 46 (1910) 217, cum syn. 
Folium lunatum minus Rumph. Herb. Amb. 5: 40, t. 25, f. 1. 
Amboina, Hoenoet, Robinson PI. Rumph. Amb. 486, October 18, 1913, 
climbing on trees, altitude 200 meters, locally known as binkuang. 

* Jackson, B. D. Index to the Linnean Herbarium (1912) 108. 
f Conard, H. S. The water lilies, a monograph of the genus Nymphaea. 
Carnegie Inst. Publ. 4 (1905) 194. 



220 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Linnaeus originally reduced Folium lunatum minus Rumph. to 
Menispermum carolinum Linn., in Stickman Herb. Amb. (1754) 
18, Amoen. Acad. 4 (1759) 128, Syst. ed. 10 (1759) 992, 
which is a wholly erroneus disposition of it. Willdenow, Sp. 
PL 4 2 (1805 ) 825, with equal error, placed it under Menispermum 
cocculus Linn. Lamarck made it the type of Menispermum 
glaucum, this species being based wholly on Rumphius's 
description and figure, so that in turn it becomes the type of 
Cocculus glaucus DC. and Pericampylus glaucus Merr. Diels 
placed Folium lunatum minus Rumph. = Cocculus glaucus. DC. 
=Menispermum glaucum Lam. as a probable synonym of Peri- 
campylus incanus (Colebr.) Miers; Engl. Pflanzenreich 46 
(1910) 217. After a careful study of the original description, 
figure, and the Amboina material cited above, I am able definitely 
to affirm that this is the correct disposition of the above names, 
but Lamarck's specific name being much the oldest is here 
adopted. 

STEPHANIA Loureiro 

STEPHANIA FORSTERI (DC.) A. Gray Bot. Wilkes U. S. Explor. Exped. 

(1854) 36. 

Cocculus forsteri DC. Syst. 1 (1818) 517. 
Convolvulus laevis III Rumph. Herb. Amb. 5: 432. 
Amboina, Paso and near the town of Amboina, Robinson PL Rumph. 
Amb. U87, July and October, 1913, in thickets along the seashore. 

The identification of Convolvulus laevis III has been made 
largely by exclusion, although Rumphius's description, as far 
as it goes, applies unmistakably to the specimen cited above. 
Moreover, the species has been previously collected in Amboina 
by Dolleschal, and extends from Java to the Philippines, northern 
Australia, and Polynesia. Hasskarl, Neue Schlussel (1866) 
143, thought that Convolvulus laevis III might be Cyclea peltata 
Hook. f. & Th.=Cyclea wallichii Diels, a species extending from 
India to the Nicobar and Andaman Islands. 

TINOSPORA Miers 

TINOSPORA RUM PHI I Boerl. in Cat. Hort. Bot. Bogor. (1901) 116, excl. 

syn. Menispermum crispum Linn, et Cocculus crispus DC. 
Funis felleus Rumph. Herb. Amb. 5: 82, t. UU, /. 1. 

This species is not represented in our Amboina collections. 
According to Rumphius's statement it was an introduced plant 
there, perhaps originating in Java. Funis felleus is cited by 
Linnaeus in the original description of Menispermum crispum 
Lmn.=Cocculus crispus T>C.=Tinospora crispa Miers; but, while 
the plate and figure are quoted correctly, the name is erroneously 



MENISPERMACEAE 



221 



cited as Funis quadrangular is, which is the name of the other 
species figured on the same plate and which is Cissus quadran- 
gularis Linn. The Linnean type was a specimen from Bengal 
and is apparently the form described by Diels as Tinospora 
crispa* The form figured by Rumphius differs notably from 
the Asiatic one in its orbicular-ovate, prominently cordate leaves, 
and Rumphius's figure agrees perfectly with material derived 
from the type plant of Boerlage's Tinospora rumphii, a specimen 
cultivated in the botanical garden at Buitenzorg, Java. I con- 
sider that Boerlage was wrong in quoting as synonyms of his 
species Menispermum crispum Linn, and Cocculus crispus DC, 
and I am disposed to disagree with Diels in his reduction here 
of Tinospora thorelii Gagnep., a cotype of which is before me. 
The authority for Tinospora crispa should manifestly be 
Miers, the combination being first published by him in Hooker 
f. and Thomson's Flora Indica 1 (1855) 183, not in Ann. Mag. 
Nat. Hist. II 7 (1851) 38, as frequently cited, and is typified 
by Menispermum crispum Linn. Menispermum tuberculatum 
Lam., to which Lamarck reduced Funis felleus Rumph. is pri- 
marily only a new name for Menispermum crispum Linn., and 
the description was based on a specimen collected by Sonnerat. 
If the Linnean species be typified by the reference to Rumphius, 
the only literature reference cited, then the name for this broad 
and prominently cordate leaved form should be Menispermum 
crispum (Linn.) Miers, but if the Linnean species be typified 
by the Bengal plant cited, then apparently Boerlage's name must 
be maintained for the Malayan form. 

Maccabuhay e Manila Rumph. Herb. Amb. 5 : 287, as to the name, 
but not the description, is a species of Tinospora, either T. reticu- 
lata Miers or the form characterized by Boerlage as T. rumphii. 
The name macabuhay is universally applied in the Philippines 
to two forms of Tinospora, but properly to the one with broad, 
deeply cordate leaves that has a very bitter principle in its 
stems. Rumphius's description of Maccabuhay e Manila ap- 
parently applies to a terrestrial orchid. 

AN AM I RT A Colebrooke 

AN AM I RTA COCCULUS (Linn.) Wight & Arn. Prodr. 1 (1834) 446. 
Menispermum cocculus Linn. Sp. PI. (1753) 340. 
Menispermum lacunosum Lam. Encycl. 4 (1797) 98 (type!). 
Cocculus lacunosus DC. Syst. 1 (1818) 519 (type!). 
Tuba baccifera Rumph. Herb. Amb. 5: 35, t. 22. 

Not represented in our Amboina collections. Linnaeus ori- 



* Engl. Pflanzenreich 46 (1910) 135. 



222 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ginally reduced Tuba baccifera Rumph. to Menispermum 
cocculus Linn., in Stickman Herb. Amb. (1754) 18, Amoen. 
Acad. 4 (1759) 128, Syst. ed. 10 (1759) 992, Sp. PL ed. 2 (1763) 
1468, which, as Anamirta cocculus (Linn.) W. & A., is certainly 
the correct disposition of it. Lamarck made it the type of 
Menispermum lacunosum, and in turn it thus became the type 
of Cocculus lacunosus DC. The species, a very characteristic 
one, is the sole known representative of the genus and extends 
from India to the Philippines and southward to New Guinea. 

ARCANGELISI A Beccari 

ARCANGEL1SIA FLAVA (Linn.) comb. nov. 

Menispermum flavum Linn, in Stickman Herb. Amb. (1754) 14, Amoen. 

Acad. 4 (1759) 128, Syst. ed. 10 (1759) 992 (type!). 
Menispermum flavescens Lam. Encycl. 4 (1797) 98 (type!). 
Anamirta flavescens Miq. Fl. Ind. Bat. 1 2 (1858) 79 (type!). 
Cocculus flavescens DC. Syst. 1 (1818) 520 (type!). 
Arcangelisia inclyta Becc. in Malesia 1 (1877) 147. 
Anamirta lemniscata Miers in Ann. Nat. Hist. Ill 14 (1864) 51. 
Arcangelisia leminscata Becc. in Malesia 1 (1877) 147; Diels in Engl. 

Pflanzenreich 46 (1910) 106, /. 38. 
Tuba flava Rumph. Herb. Amb. 5: 38, t. 24. 

This is not represented in our Amboina collections. Tuba 
flava Rumph. has been confused in recent botanical literature 
with Anamirta cocculus Wight & Arn., and several of the above 
names are cited as synonyms of Anamirta cocculus (Linn.) 
Wight & Arn. in the most recent monograph of the family; * 
Menispermum flavum Linn, is not accounted for in this mono- 
graph. The first four names cited above are based absolutely 
and wholly on Tuba flava Rumph. That Arcangelisia and not 
Anamirta is the proper disposition of it is proved by the original 
description, leaves 5-nerved at the base, "autem cordiformia 
sed inf erius non excavata ;" by the fruit, size, and other charac- 
ters; and especially by the color of the woody tissue, "interne 
eleganter flavescunt;" this last character alone being one by 
which Arcangelisia can always be distinguished from Anamirta. 
Loureiro, Fl. Cochinch. (1790) 626, cites Tuba flava Rumph. as 
a synonym of Fibraurea tinctoria Lour., a valid species, the 
type of the genus Fibraura, but one that has little in common 
with Arcangelisia. Arcangelisia flava extends from Java to 
Luzon and New Guinea, and was collected in Amboina by 
Teysmann. 



* Diels in Engl. Pflanzenreich 46 (1910) 108. 



MAGNOLIACEAE 



223 



MAGNOLIACEAE 

MICH ELI A Linnaeus 

MICHELIA CHAMPACA Linn. Sp. PL (1753) 536. 
Michelia suaveolens Pers. Syn. 2 (1807) 94. 
Michelia parviflora DC. Syst. 1 (1818) 449? (type!). 
Michelia caerulea DC. Syst. 1 (1818) 449? (type!). 
Michelia blumei Steud. Nomencl. ed. 2, 2 (1841) 139. 
Michelia euonymoides Burm. f. Fl. Ind. (1768) 124, p. p. 
Sampacca domestica Rumph. Herb. Amb. 2: 199, t. 67. 
Sampacca II parviflora Rumph. Herb. Amb. 2: 200? 
Sampacca Ml coerulea Rumph. Herb. Amb. 2: 200? 

The common champaca is not represented in our Amboina 
collections. Sampacca domestica was described and figured by 
Rumphius from cultivated specimens and is certainly typical 
Michelia champaca Linn. It was first reduced by Linnaeus, in 
Stickman Herb. Amb. (1754) 10, Amoen. Acad. 4 (1759) 121; 
while in the Systema, ed. 10 (1759) 1082, Sampacca silvestris 
Rumph., 1. c. t. 68, is added ; see below under Michelia tsiampacca 
Linn. Michelia suaveolens Pers. is merely a new name for M. 
champaca Linn., while M. blumei Steud. was proposed as a new 
name for Michelia champaca Blume, which, however, is typical 
M. champaca Linn. The doubtful synonyms mentioned above, 
both of de Candolle and of Rumphius, must await further 
exploration of the Malayan region before they can be definitely 
placed. Michelia parviflora DC. was based wholly on Sampacca 
parviflora Rumph., which is very briefly described by Rumphius 
from Ternate specimens as being similar to his Sampacca 
domestica, but with smaller flowers ; it is probably merely a form 
of M. champaca Linn. Michelia caerulea DC. was based wholly 
on Sampacca coerulea Rumph., also very briefly described from 
Javan specimens similar to Sampacca domestica, but with bluish 
flowers; the Javan name cited is tsjampacca biru, and it is 
probably merely a form of the common Michelia champaca Linn. 

MICHELIA ALBA DC. Syst. 1 (1818) 449 (type!). 
Michelia longifolia Blume Bijdr. (1825) 7. 
Sampacca domestica IV alba Rumph. Herb. Amb. 2: 200. 

Michelia alba DC. was based wholly on Sampacca alba Rumph,, 
which in turn was described from the white-flowered cultivated 
form known in Java as tsjampacca puti. Michelia longifolia 
Blume is unquestionably the same form, but de Candolle's name 
is the older and is here retained. 



224 



RUMPH IUS'S HERBARIUM AMBOINENSE 



MICH EL I A TSIAMPACCA Linn. Mant. 1 (1767) 78 (type!). 
Sampacca silvestris Rumph. Herb. Amb. 2: 202. t. 68. 

This species is not represented in our Amboina collections. 
It was described by Rumphius from sylvan specimens growing 
in Amboina, having white flowers. It is probably allied to 
Michelia montana Blume. The Rumphian reference seems to be 
the basis of Michelia tsiampacca Linn., but the plate was origin- 
ally referred by Linnaeus, Syst. ed. 10 (1759) 1082, to Michelia 
champaca Linn. By other authors it has been referred to 
Michelia suaveolens Pers., M. sericea Pers., and M. euonymoides 
Burm. f ., all of which appear to be synonyms of Michelia cham- 
paca Linn. The native Amboinese name cited by Rumphius is 
tsjampacca utan or tsjampacca puti, and botanical material from 
Amboina will be necessary before the exact status of the species 
can be determined. 

TALAUMA Jussieu 

TALAUMA RUMPHII Blume Bijdr. (1825) 10 (type!). 

Liriodendron liliifera Linn. Sp. PI. ed. 2 (1762) 755 (type!) non 

Talauma liliifera Kurz. 
Sampacca montana Rumph. Herb. Amb. 2: 204, t. 69 haud Arbor 

violaria Rumph. 1. c. 203! 

Not represented in our Amboina collections, but manifestly 
a Talauma. Sampacca montana Rumph. is the whole basis of 
Liriodendron -liliifera Linn., but Talauma liliifera Kurz was 
based on Liriodendron liliifera Roxb., non Linn., so that the 
Linnean specific name is invalid in Talauma. The Rumphian 
plant is also the whole basis of Talauma rumphii Blume, but 
the species, as yet not represented by any botanical material 
definitely known to represent the Rumphian plant, is one of 
doubtful status. Sampacca montana was referred, with doubt, 
to Magnolia pumila Andr. by de Candolle, Syst. 1 (1818) 458, 
and it certainly is not this species, nor is it Magnolia inodora 
DC. 1. c. 459, which was based on Liriodendron liliifera Lour., 
Fl. Cochinch. (1790) 346; Loureiro's species was based! on 
specimens from Canton, China, to which he added a reference 
to Sampacca montana Rumph. 

Arbor violaria Rumph., Herb. Amb. 2: 203, as described, is 
entirely different from Sampacca montana Rumph. h c. 204, 
t. 69, as described and figured; the plate goes with Sampacca 
montana, not with Arbor violaria Rumph. It has been assumed 
by some that the descriptions applied to the same plant, but 
Arbor violaria was described from cultivated specimens from 
Banda Island, while Sampacca montana was described from 



ANNONACEAE 



225 



sylvan specimens collected in Amboina. The descriptions apply 
to totally different plants, and I am unable to suggest the proper 
position of Arbor violaria. 

ANNONACEAE 

UVARIA Linnaeus 
UVARIA MUSARIA (Dunal) DC. Mem. Anon. (1832) 29 (type!). 
Unona musaria Dunal Monogr. Anon. (1817) 100 (type!). 
Uvaria moluccana Kostel. Allg. Med.-Pharm. Fl. 5 (1836) 1707 
(type!). 

Funis musarius latifolius Rumph. Herb. Amb. 5: 78, t. 1+2. 

Amboina, Amahoesoe and Hitoe messen, Robinson PI. Rumph. Amb. 79, 
September and October, 1913, in flower, growing in forests, altitude 60 to 
200 meters; Liang, Robinson PL Rumph. Amb. 1+79, November 29, 1913, 
in thickets at sea level, with nearly full-grown but immature fruits. 

Uvaria musaria (Dunal) DC. has been previously known only 
from the Rumphian figure and description, this being the whole 
basis of Unona musaria Dunal, Uvaria musaria DC, and Uvaria 
moluccana Kostel. In vegetative and floral characters it closely 
approximates Uvaria rosenhergiana Scheff., of New Guinea. 
Burman f., Fl. Ind. (1768) 124, followed by Lamarck, Willdenow, 
Persoon, Poiret, and Pritzel, erroneously reduced Funis musarius 
latifolius Rumph. to Uvaria zeylanica Linn.; Blume, Fl. Jav. 
1 (1828) Anon. 22, equally in error, placed it under Uvaria 
hirsuta Blume; and Wight and Arnott, Prodr. 1 (1834) 9, placed 
it under Uvaria macrophylla Roxb. Uvaria musaria (Dunal) 
DC. seems to be a perfectly valid species, in the alliance with 
Uvaria rosenhergiana Scheff., U. littoralis Blume, and U. ovali- 
folia Blume, differing radically, however, in its slightly pubescent 
leaves and in its elongated puberulent fruits. Uvaria pilosa 
Roxb., type from the Moluccas, should be critically compared. 

UVARIA sp. 

Funis musarius angustifolius Rumph. Herb. Amb. 5: 78. 

The exact status of this form is indeterminable without ma- 
terial from Amboina. Dunal, Monog. Anon. (1817) 99, reduced 
it, with doubt, to Unona narum Dunal, a species typified by 
Narum-panel Rheede, Hort. Malabar. 2: 11, t. 9. Blume, Fl. 
Jav. 1 (1828) Anon. 24, thought that it might be Uvaria ar gen- 
tea Blume, while Wight and Arnott, Prodr. 1 (1834) 9, placed it, 
with doubt, under Uvaria grandiflora Roxb. 

UVARIA sp. 

Funis dentarius Rumph. Herb. Amb. 5: 79. 

Like the preceding, this cannot be definitely placed within 

144971 15 



226 



RUMPHIUS'S HERBARIUM AMBOINENSE 



the genus Uvaria without additional material from Amboina. 
It is undoubtedly a species of Uvaria, and it has been suggested 
as a possible synonym of Uvaria littoralis Blume or U. latifolia 
Blume; see Hasskarl, Neue Schlussel (1866) 97. The form very 
briefly mentioned by Rumphius as Funis dentarius niger, Herb. 
Amb. 5: 79, represents either the same species as Funis dentarius 
Rumph. or a closely allied one. 

CANANGIUM Baillon 

CANANGI U M ODORATUM (Lam.) Baill. ex King in Journ. As. Soc. 
Beng. 61 2 (1892) 41. 

Uvaria odorata Lam. Encycl. 1 (1785) 595. 
Unona odorata Dunal Monog. Anon. (1817) 108. 
Cananga odorata Hook. f. & Th. Fl. Ind. 1 (1855) 130. 
Cananga Rumph. Herb. Amb. 2: 195, t. 65. 

Amboina, Elephant River, Robinson PI. Rumph. Amb. 80, July 19, 1913, 
locally known as bunga kanangan. 

Lamarck's original description was primarily based on speci- 
mens collected by Sonnerat, Cananga Rumph. being reduced as 
a synonym. In this reduction, as Canangium odoratum Baill., 
certainly the correct disposition of it, he was followed by 
Willdenow, Persoon, Blume, Spanoghe, and Roxburgh. Dunal 
cites Cananga as a synonym of Unona odorata Dun. in transfer- 
ring the species to that genus. Cananga Hook, f & Thomson 
(1855), the generic name from Rumphius, is invalidated by 
Cananga Aubl. (1775), for which reason Baillon has proposed 
the new generic name Canangium, Hist. PL 1 (1868) 213, but 
Baillon did not actually transfer the species to this genus, this 
being apparently first accomplished by King. In the more recent 
literature Cananga Rumph. is generally cited under the name 
Canangium odoratum Baill., in the somewhat older literature, 
after the year 1855, under Cananga odorata Hook. f. & Th. 
Pritzel and Burman f . are wrong in referring it to Uvaria zey- 
lanica Linn., which is a true Uvaria, known only from India and 
Ceylon. 

POLYALTHIA Blume 

POLYALTHIA sp. 

Uvaria ligularis Lam. Encycl. 1 (1785) 597 (type!). 
Unona ligularis Dunal Monog. Anon. (1817) 110 (type!). 
Cananga silvestris II angustifolia Rumph. Herb. Amb. 2: 197, t. 66, 

f. 2. 

Nothing remotely resembling this occurs in our Amboina collec- 
tions. The figure and the description apply very closely to 



ANNONACEAE 



227 



Polyalthia lateriflora (Blume) King, a species originally de- 
scribed from Javan material, now known from the Malay 
Peninsula and Java, and the very closely allied Polyalthia 
zamboangensis Merr., of Mindanao. Cananga silvestris II angus- 
lifolia Rumph. is the whole basis of Uvaria ligularis Lam. and 
of Unona ligularis Dunal, and Lamarck's specific name may 
prove to be the oldest for the plant now known as Polyalthia 
lateriflora King. In the absence of material from Amboina re- 
presenting Rumphius's species, however, it is considered advis- 
able to defer the actual transfer of Uvaria ligularis Lam. to 
Polyalthia, although the species manifestly belongs in the latter 
genus. Linnaeus reduced it, by error, to Uvaria zeylanica Linn., 
in Stickman Herb. Amb. (1754) 10, Amoen. Acad. 4 (1759) 
121, including also Cananga silvestris I trifolia, which is figured 
on the same plate, but neither of which remotely resembles 
Uvaria zeylanica Linn., a species known only from India and 
Ceylon. 

POLYALTHIA sp. 

Guatteria rumphii Blume ex Henschel Vita Rumph. (1833) 153 
(type!). 

Arbor nigra parvifolia Rumph. Herb. Amb. 3: 10, 11, t. U, f- 2; t. 5. 

The plant figured and described is certainly a Polyalthia, but 
its status cannot be definitely settled without material from Am- 
boina. It is the whole basis of Guatteria rumphii Blume as 
published by Henschel, by citation of Rumphius, as indicated 
above, a name that has been entirely overlooked, and one that 
is not included in Index Kewensis. Linnaeus reduced, with 
doubt, both t. U and t. 5 to Uvaria zeylanica Linn., in Stickman 
Herb. Amb. (1754) 11, Amoen. Acad. 4 (1759) 122, but neither 
figure presents anything in common with this species. Hasskarl, 
Neue Schliissel (1866) 46, quotes Teysmann's opinion that Arbor . 
nigra parvifolia Rumph. represented Artabotrys suaveolens 
Blume, a species with which neither the description nor the 
figure agrees. 

Arbor nigra maculosa Rumph. Herb. Amb. 3: 12, t. U, f. 1, is 
probably one of the Annonaceae, possibly a species of Polyalthia. 
It is certainly not Artabotrys odoratissimus Blume, where it 
was placed by Henschel, Vita Rumph. (1833) 153. 

Arbor nigra latifolia Rumph. Herb. Amb. 3 : 12, is probably some 
annonaceous plant, but its status must await a more exhaustive 
botanical exploration of Amboina. No reduction of it has ever 
been suggested. 



228 



RUMPHIUS'S HERBARIUM AMBOINENSE 



GON IOTHALAM US Hooker f. and Thomson 

GON 10TH ALAM US sp. 

Uvaria tripetala Lam. Encycl. 1 (1785) 597 (type!). 

Unona tripetaloidea Dunal Monog. Anon. (1817) 104 (type!). 

Unona tripetala DC. Prodr. 1 (1824) 90 (type!). 

Cananga silvestris I trifolia Rumph. Herb. Amb. 2: 197, t. 66, f. 1. 

This is not represented in our Amboina collections. The 
species, as described and figured by Rumphius, is apparently, 
a very characteristic one, and Amboina material representing 
it, when collected, should be connected with it with little dif - 
ficulty. Cananga silvestris I trifolia Rumph. is the whole basis 
of the three names Uvaria. tripetala Lam., Unona tripetaloidea 
Dunal, and U. tripetala DC. Poiret, in Lam. Encycl. 8 (1808) 
187, referred it to Unona discolor Dunal=Desmos chinensis 
Lour., an impossible reduction; while Hasskarl, Neue Schlussel 
(1866) 40, thought that it might be a species of Artabotrys, 
also an impossible reduction. As the three inner petals are de- 
scribed as much smaller than the outer three and as surrounding 
or covering the stamens, it is probable that Goniothalamus is its 
correct genus, as here tentatively suggested. However, no 
actual transfer is here made, as the exact status of the species 
must await further botanical exploration of Amboina. 

ARTABOTRYS R. Brown 

ARTABOTRYS SUAVEOLENS Blume Fl. Jav. 1 (1828) Anon. 62. 
Spina vaccarum Rumph. Herb. Amb. 5: 21, t. 11+. 

This species is not represented in our Amboina collections. 
Spina vaccarum Rumph. was reduced by Blume to Artabotrys 
suaveolens Blume in the original description of that species, 
and the figure is apparently a good representation of it as 
currently interpreted. All authors have followed Blume in this 
reduction, the species being one of very wide distribution in 
the Malayan region. When Amboina material is available for 
study, however, it should be critically compared with Artabotrys 
inodorus Zipp. of New Guinea. 

MELODORUM Hooker f. and Thomson 

MELODORUM LATI FOLIUM (Dunal) Hook. f. & Th. Fl. Ind. (1855) 115, 
saltern quoad syn. 
Unona latifolia Dunal Monog. Anon. (1817) 115 (type!). 
Uvaria latifolia Blume Fl. Jav. 1 (1828) Anon. 37. 
Cananga silvestris III latifolia Rumph. Herb. Amb. 2: 198. 

This is not represented in our Amboina collections. Unona 
latifolia Dunal, as originally published, was based wholly on 



MYRISTICACEAE 



229 



Rumphius's description of Cananga silvestris III latifolia, and 
Blume redescribed it from Javan material as Uvaria latifolia 
Blume, followed by Hooker f. and Thomson, Miquel, and King, 
as Melodorum lati folium (Dunal) Hook. f. & Th. It is probable 
that Melodorum latifolium Hook. f. & Th., as described in 
modern botanical literature, is the same as the form described by 
Rumphius, the original basis of the species; yet no botanical 
material from Amboina representing the species as it is at 
present understood seems to be extant, and the exact status of 
Unona latifolia Dunal must await further botanical exploration 
of Amboina. 

AN NONA Linnaeus 

ANNONA RETICULATA Linn. Sp. PL (1753) 537. 
Anona Rumph. Herb. Amb. 1 : 136, t. J^5. 

This commonly cultivated fruit tree is not represented in our 
Amboina collections, but like the next, it doubtless still occurs 
there, as both are widely distributed in the Malayan region. 
Anona Rumph. was first reduced to Annona reticulata Linn., 
in Stickman Herb. Amb. (1754) 7, Amoen. Acad. 4 (1759) 119, 
Syst. ed. 10 (1759) 1083, Sp. PI. ed. 2 (1763) 757, which is 
certainly the correct disposition of it. De Candolle, Syst. 1 
(1818) 474, placed it under Annona mucosa Aub\.= Rollinia mu- 
cosa Baill., in this erroneous reduction being followed by Don, 
Henschel, Dietrich, and Pritzel. 

ANNONA SQUAMOSA Linn. Sp. PL (1753) 537. 

Anona tuberosa Rumph. Herb. Amb. 1: 138, t. 4.6. 

Like the above, this commonly cultivated plant is not repre- 
sented in our Amboina collections. Anona tuberosa Rumph. 
was first reduced by Linnaeus, in Stickman Herb. Amb. (1754) 
7, Amoen. Acad. 4 (1759) 119; this reduction has been con- 
sistently followed by most authors and is certainly the correct 
disposition of the species. 

MYRISTICACEAE 
MYRISTICA * Linnaeus 

MYRISTICA FRAGRANS Houtt. Handleid. 2 3 (1774) 333. 

Myristica officinalis Linn. f. Suppl. (1781) 265. 

Myristica moschata Thunb. Act. Holm. (1782) 45. 

Myristica aromatica Lam. Act. Paris (1788) 155. 

Nux myristica Rumph. Herb. Amb. 2: 14, t. 4. 
Amboina, Kati-kati and Way tommo, Robinson PL Rumph. Amb. 245, 
246, August and October, 1913, from cultivated plants, altitude 50 to 60 
meters, locally known as pala. 



* Retained name, Vienna Code; Quret Adans. (1763) is older. 



230 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Mux myristica Rumph. has been cited by various authors under 
all of the names listed above. Rumphius described five forms, 
all apparently merely variants of the common nutmeg, although 
Miquel referred Pala radja Rumph. to Myristica radja Miq., 
Ann. Mus. Bot. Lugd. Bat. 1 (1864) 206. Warburg, Nov. Act. 
Akad. Naturf. 68 (1897) 453, considers the plant that Miquel 
described to be Myristica speciosa Warb., not M. fragrans 
Houtt. 

MYRISTICA FATUA Houtt. Handleid 2 3 (1774) 337. 
Myristica tomentosa Thunb. Act. Holm. (1782) 46. 
Myristica spadicea Blume Bijdr. (1825) 577. 
Myristica macrophylla Roxb. Fl. Ind. ed. 2, 3 (1832) 846. 
Nux myristica mas Rumph. Herb. Amb. 2: 24, t. 5. 

Nothing resembling Myristica fatua Houtt., as currently 
interpreted, occurs in our Amboina collections. Nux myristica 
mas has been referred by various authors to Myristica fatua 
Houtt., M. tomentosa Thunb., M. officinalis Gaertn., M. philip- 
pensis Lam., M. malabarica Lam., and M. moschata Thunb., the 
reductions in all but the two cases first cited being erroneous. 
The species is a well-known one, occurring in Banda, Amboina, 
Tidore, and ( ?) Buru, and is cultivated in the botanic garden 
at Buitenzorg, Java. 

HORSFI ELDI A Willdenow 

HORSFI ELDI A SYLVESTRIS (Houtt.) Warb. in Nov. Act. Acad. Naturf. 
68 (1897) 337, t. 12, f. 1-6. 
Myristica sylvestris Houtt. Handl. 2 3 (1774) 326. 
Myristica salicifolia Willd. Sp. PI. 4 (1805) 871. 

Myristica pinnae formis Miq. Ann. Mus. Bot. Lugd. Bat. 2 (1865) 49. 
Palala secunda Rumph. Herb. Amb. 2: 26, t. 6 (poor). 
Amboina, Hitoe messen, Robinson PI. Rumph Amb. 235, November 1, 
1913, in forested ravines, altitude about 100 meters. 

The Rumphian figure and description are perhaps the type of 
Myristica sylvestris Houtt. (original publication not seen by me) ; 
Myristica salicifolia Willd. is merely a new name for M . sylves- 
tris Houtt. 

HORSFI ELDI A CAN ARI FORM IS (Blume) comb. nov. 

Myristica canariformis Blume Rumphia 1 (1835) 190 (type!). 
Horsfieldia roxburghii Warb. in Nov. Act. Akad. Naturf. 68 (1897) 

277, t. 21, f. 1-2. 
Palala quarta Rumph. Herb. Amb. 2: 27, t. 8. 
Amboina, Hitoe messen, Robinson PI. Rumph. Amb. 2U0, November 1, 
1913, in forests, altitude about 175 meters. 



MYRISTICACEAE 



231 



Palala quarta Rumph. is the whole basis of Myristica canari- 
formis Blume, which Warburg thought was a possible synonym 
of Horsfielclia nesophila (Miq.) Warb. I am of the opinion, 
however, that it is identical with the Amboina form described by 
Warburg as Horsfieldia roxburghii and accordingly have here 
adopted Blume's specific name. Willdenow, Sp. PL 4 (1805) 
871, considered that it represented a variety of Myristica micro- 
carpa Willd., whatever that species may be, but it is certainly 
not properly placed here. 

HORSFIELDIA sp. 

Myristica tingens Blume Rumphia 1 (1835) 190 (type!). 
Palala tertia Rumph. Herb. Amb. 2: 27, t. 7. 

According to Rumphius's description and figure this is a very 
characteristic species, but nothing in our Amboina collections 
can be referred to it. Palala tertia Rumph. is the whole basis of 
Myristica tingens Blume, but no new combination is here made 
in view of the uncertain status of that species. Willdenow, 
Sp. PI. 4 (1805) 871, referred it to Myristica microcarpa Willd., 
which is perhaps the oldest valid specific name for the species. 
Lamarck, Encycl. 4 (1797) 391, placed it with doubt under 
Myristica uviformis Lam., but Lamarck's species is not a myris- 
ticaceous plant, and Warburg has suggested that it belongs in 
the Euphorbiaceae. The status of Myristica microcarpa Willd. 
is entirely doubtful, for I cannot agree with Warburg that it 
is a synonym of Knema cinerea Warb. Incidentally the type of 
Knema cinerea (Poir.) Warb. certainly did not come from the 
Philippines as Warburg states, for LaBillardiere never visited 
the Archipelago. Buton, as spelled by him, and Bouton, as 
spelled by Poiret in the original description, probably refer to 
Boeton Island, near the southern end of Celebes. 

HORSFIELDIA sp. 

Myristica aruana Blume Rumphia 1 (1835) 191 (type!). 
Palala aruana Rumph. Herb. Amb. 7: 56, t. 24, f. 3. 

A species of doubtful status, probably, however, a Horsfieldia. 
Myristica aruana Blume was based wholly on Palala aruana 
Rumph. and accordingly must be interpreted solely from Rumph- 
ius's description and crude figure. If Warburg is correct in 
reducing Myristica aruana Blume to Horsfieldia novo-guineensis 
Warb., then Blume's specific name should be adopted for the 
species unless one considers it invalidated by the distinct Hors- 
fieldia aruensis Warb. 



232 



RUMPHIUS'S HERBARIUM AMBOINENSE 



GYMNACRANTHERA Warburg 

GYMNACRANTHERA ZIPPELIANA (Miq.) Warb. in Nov. Act. Akad. 
Naturf. 68 (1897) 372. 

Myristica zippeliana Miq. Ann. Mus. Bot. Lugd. Bat. 2 (1865) 50. 
Palala quinta Rumph. Herb. Amb. 2: 28, t. 9. 
Amboina, Hitoe messen, Robinson PL Rumph. Amb. 239, October 18, 
1913, in forests, altitude about 450 meters, locally known as palala utan. 

Willdenow, Sp. PL 4 (1805) 871, considered Palala quinta 
Rumph. to be a variety of Myristica microcarpa Willd., while 
Lamarck, Encycl. 4 (1797) 388, placed it with doubt under 
Myristica globularia Lam. Lamarck's species was based on 
specimens collected by Sonnerat and is Knema globularia Warb. 
There is little doubt that the specimen cited above represents 
Palala quinta Rumph., and it is certainly a Gymnacr anther a, 
probably G. zippeliana Warb., although this species has not been 
previously reported from Amboina. 

KNEMA Loureiro 

KNEMA TOMENTELLA Warb. in Nov. Act. Akad. Naturf. 68 (1897) 588, 
t. 25, f. 1-2. 
Palala sexta Rumph. Herb. Amb. 2: 28. 
Amboina, Hatiwe, Wae, and Kati-kati, Robinson PI. Rumph. Amb. 236, 
237, 238, September to November, 1913, in forests, altitude 20 to 350 
meters, locally known as palala utan. 

Palala sexta Rumph. is, with little doubt, the same as Knema 
tomentella Warb. Hasskarl, Neue Schliissel (1866) 26, sug- 
gested that it might be the same as Myristica corticosa Hook. f. 
& Th., which Warburg interprets as being composed of Knema 
angustifolia Warb., K. glauca Warb., and K. missionis Warb. 

LAURACEAE 

CINNAMOMUM Linnaeus 

CINNAMOMUM CULILAWAN Blume Bijdr. (1825) 571, Rumphia 1 (1835) 
26, t. 9, f. 1. 

Laurus culitlawan Linn, in Stickman Herb. Amb. (1754) 9, Amoen. 

Acad. 4 (1759) 120 (type!). 
Laurus culilaban Linn. Mant. 2 (1771) 237 (type!). 
Cortex caryophylloides albus Rumph. Herb. Amb. 2: 65, t. lb. 
Culit La wan Rumph. Herb. Amb. 7: 65. 

The Rumphian figure and description of Cortex caryophy- 
lloides albus are the whole basis of both Laurus culitlawan Linn, 
and L. culilaban Linn., but Cinnamomum culilawan Blume was 
published independently of the Linnean binomials. Blume gives 
an ample description and figures from Amboina material; 



LAURACEAE 



233 



however, he definitely excludes the figure of the inflorescence as 
given by Rumphius, which he apparently thought went with 
Cinnamomum caryophylloides ruber Rumph. 

CINNAMOMUM CU LI LA WAN Blume var. RUBRUM (Blume) Meissn. in 
DC. Prodr. 15 1 (1864) 14. 
Cinnamomum rubrum Blume Rumphia 1 (1835) 29. 
Cinnamomum caryophylloides ruber Rumph. Herb. Amb. 2: 66. 

Blume based his description of Cinnamomum rubrum partly on 
the form described by Rumphius above cited, which Meisner 
considers to be a variety of Cinnamomum culilawan Blume. 

CINNAMOMUM XANTHONEURUM Blume Rumphia 1 (1825) 33. 

Culitlawan ex Papuanis et Moluccis insulis Rumph. Herb. Amb. 2: 66. 

This follows Blume's reduction of the form Rumphius de- 
scribed, which is probably the correct disposition of it. 

CINNAMOMUM JAVANICUM Blume Rumphia 1 (1835) 42. 
Sin doc Rumph. Herb. Amb. 2: 69. 

The reduction of Sindoc follows Blume's disposition of it. The 
form Rumphius described was placed by Burman f . under Laurus 
malabathrum Burm. f., Fl. Ind. (1768) 92, based in part on 
a species figured and described by Rheede, and in part on Sindoc 
of Rumphius, to be typified by the former. Miquel placed it 
under Cinnamomum sulphuratum Nees ; Henschel placed it under 
Cinnamomum sintoc Blume; and Nees followed Blume in re- 
ducing it to Cinnamomum javanicum Blume. 

CINNAMOMUM CAMPHORA (Linn.) T. Nees & Eberm. Handb. Med.- 
Pharm. Bot. 2 (1831) 430. 
Laurus camphora Linn. Sp. PI. (1753) 369. 
Arbor camphorifera I vera Rumph. Herb. Amb. 7: 65, 68. 

The general discussion is manifestly in part applicable to 
the true camphor tree, Cinnamomum camphora T. Nees & Eberm. 

CINNAMOMUM spp. indet. 

Laurus japan ica Rumph. Herb. Amb. 7: 64. 
Cinnamomum japanicum II Rumph. Herb. Amb. 7: 64. 
Cinnamomum zeylanicum Rumph. Herb. Amb. 7: 64. 

These three forms, imperfectly described, are manifestly ref- 
erable to the genus Cinnamomum, but it is impossible to deter- 
mine just which species were intended. The first was placed by 
Henschel under Laurus soncaurium Ham.= Cinnamomum tamala 
T. Nees & Eberm., of the Himalayan region, certainly a wrong 
disposition of it; the second was thought by Hasskarl, Neue 
Schlussel (1866) 192, to be possibly referable to Cinnamomum 
dulce Nees ; and the third was placed by Hasskarl under Cinna- 



234 



RUMPHIUS'S HERBARIUM AMBOINENSE 



momum zeylanicum Nees, which is perhaps the correct disposi- 
tion of it. 

DEHAASIA Blume 

DEHAASIA MEDIA Blume Rumphia 1 (1835) 163, t. 45. 
Haasia media Nees Syst. Laur. (1836) 375. 
Machilus Hi media Rumph. Herb. Arab. 3: 70, t. 41. 

This is not represented in our Amboina collections. Blume's 
ample description and figure were based on Amboina material, 
and his disposition of Machilus media Rumph. is doubtless cor- 
rect, Rumphius's description and figure being cited in the orig- 
inal description of the species. Nees, Syst. Laur. (1836) 125, 
thought that it might be Persea peduncularis Nees, and it was 
cited by Henschel as Machilus peduncularis Nees. 

MACHILUS Nees 

MACHILUS sp.? 

Machilus IV minima Rumph. Herb. Arab. 3: 70, t. 42. 

Nothing resembling the plant figured and described is pre- 
sented by our Amboina collections, but a species of Machilus, 
as interpreted by Meisner, is probably intended. The generic 
name Machilus of Nees is taken from Rumphius. Loureiro, 
Fl. Cochinch. (1790) 253, referred Machilus minima with doubt 
to Laurus indica Linn., an entirely wrong disposition of it. Nees 
thought that the figure represented Machilus odoratissima Nees, 
Syst. Laur. (1836) 172, and repeated the reduction in DC. Prodr. 
15 1 (1864) 40. It is certainly not Machilus odoratissima Nees 
and may ultimately prove to belong to some other genus. It is 
possibly a species of Phoebe rather than of Machilus. 

EUSI DEROXYLON Teysmann and Binnindyck 

EUSI DEROXYLON ZWAGERI Teysm. & Binn. in Nat. Tijdschr. Nederl. 
Ind. 25 (1863) 292. 
Lontaro simile lignum Rumph. Herb. Amb. 1: 52. 

Rumphius includes only a brief description of the Bornean 
wood which he calls caju boelian; the identification has been 
made from the native name cited, billian being the common name 
for this important Bornean timber tree. 

LITSEA* Lamarck 
LITSEA RUMPHM (Blume) F.-Vill. Noviss. App. Fl. Filip. (1880) 180. 
Tetranthera rumphii Blume Mus. Bot. 1 (1851) 382. 
Lignum leve alterum Rumph. Herb. Amb. 3: 72, t. 45. 

* Retained name, Vienna Code; Malapoenna Adans. (1763) and Tomex 
Thumb. (1783) are older. 



LAURACEAE 



235 



Not represented in our Amboina collections. The reduction 
was made by Blume in the original description of Tetranthera 
rumphii Blume, and this is presumably the correct disposition 
of the Rumphian species ; Blume apparently had a specimen from 
Amboina, judging from his short and imperfect description. 
Nees thought that it was a species of Tetranthera near T. mono- 
petala Roxb. 

LITSEA sp. 

Glabraria tersa Linn. Mant. 2 (1771) 276, quoad syn. Rumph. 
Lignum leve angustifolium Rumph. Herb. Amb. 3: 71, t. UU. 

Not represented in our Amboina collections. The plant that 
Rumphius figured and described is manifestly a species of Litsea, 
apparently in the group with Litsea fulva F.-Vill. and Litsea 
luzonica F.-Vill. Blume, Mus. Bot. 1 (1851) 383, placed it 
as a possible synonym of Tetranthera forstenii Blume =Litsea 
forstenii Boerl. Its exact status, however, cannot be determined 
without Amboina material. Linnaeus, Mant. 2 (1771) 276, 
quoted Lignum leve angustifolium Rumph. as a synonym of 
Litsea tersa Linn, in the original description of that species. 
It is clearly manifest, however, from the description, that he had 
an actual specimen, and it is equally manifest that the plant he 
described is not the same as the one figured and described by 
Rumphius. The specimen in the Linnean herbarium has been 
examined for me by Mr. Gamble, who writes under date of June 
17, 1917, that it is a polypetalous plant, probably belonging 
in the Bombacaceae near the genus Boschia Korth. Glabraria 
Linnaeus must, accordingly, be eliminated as a synonym of 
Litsea. Persoon, Syn. 2 (1807) 4, following Linnaeus's errone- 
ous reduction of Lignum, leve angustifolium, quotes it as a 
synonym of Litsea glabraria Pers., as does Nees under Tetran- 
thera glabraria Nees ; these names, however, go with Litsea gluti- 
nosa (Lour.) C. B. Rob. (L. chinensis Lam., L. sebifera Pers.), 
as does Litsea tersa Merr. (non Glabraria tersa Linn.). 

LITSEA STICKMANII sp. nov. 

Lignum leve latifolium Rumph. Herb. Amb. 3: 71, t. J>3. 
Amboina, Hitoe messen and Paso, Robinson PL Rumph. Amb. U70, U71 
(type), November 1 and 25, 1913, in forests and along roadsides, altitude 
3 to 200 meters, locally known as halaor pantey. 

Arbor 10 ad 12 m alta, inflorescentiis exceptis glabra; foliis 
oblongis, firme chartaceis, usque ad 20 cm longis, utrinque 
aequaliter angustatis, basi acutis, apice acutis vel obscure acu- 
minatis, supra subolivaceis, subtus pallidis, nervis utrinque 
circiter 8, subtus prominentibus, curvato-adscendentibus, re- 
ticulis distinctis; inflorescentiis axillaribus, fasciculatis, umbel- 



236 



RUMPHIUS'S HERBARIUM AMBOINENSE 



lulis circiter 6-floris, longe graciliterque pedunculatis, floribus 
pubescentibus. 

A tree 10 to 12 m high, quite glabrous except the inflorescence. 
Branches and branchlets brownish, the latter more or less angled. 
Leaves alternate, oblong, firmly chartaceous, 12 to 20 cm long, 
3.5 to 5.5 cm wide, equally narrowed to the acute base and 
to the acute or obscurely acuminate apex, the upper surface more 
or less olivaceous when dry, smooth and shining, the lower pale ; 
lateral nerves about 8 on each side of the midrib, prominent on 
the lower surface, curved-ascending, obscurely anastomosing, 
the ultimate reticulations fine, rather close; petioles about 1 
cm long. Flowers axillary, the umbellules fascicled, their pe- 
duncles slender, slightly pubescent, about 1 cm long, each about 
6-flowered. Bracts obovate to elliptic-obovate, rounded, con- 
cave, pubescent, 3.5 to 4 mm long. Flowers appressed-pubes- 
cent, their pedicels 2 to 3 mm long, the lobes oblong, 2 mm in 
length. Fruit unknown, when very young one to three on' 
each peduncle. 

This species is apparently allied to the form Blume described 
as Tetranthera ambigua Blume (not Litsea ambigua Nees), but 
differs in its smaller leaves, glabrous branchlets, and other 
characters. It certainly represents Lignum leve latifolium 
Rumph. It is dedicated to 0. Stickman, author of the first 
publication on the Herbarium Amboinense. 

Loureiro, Fl. Cochinch. (1790) 471, erroneously referred Lig- 
num leve latifolium Rumph. to Glabraria tersa Linn., followed 
by Blume's reference of it to Tetranthera laurifolia Jacq. var. 
tersa (Linn.) Blume, with which it has nothing in common. 
Nees, in DC. Prodr. 15 1 (1864) 180, placed it, with doubt, under 
Tetranthera laurifolia Jacq. var. saligna Nees, where it certainly 
does not belong. 

LITSEA sp. 

Machilus II femina Rumph. Herb. Amb. 3: 69, t. 40, f. B. 
Amboina, Hoetoemoeri road, Robinson PI. Rumph. Amb. 469, September 
30, 1913, in forests, altitude about 300 meters. 

The specimen cited certainly represents the form described 
by Rumphius, of which he figures a single leaf. There is nothing 
in the figure by which Machilus I mas can be distinguished from 
Machilus femina, and it certainly represents a species of Litsea 
perhaps not distinct from the one here considered. The only 
previous suggestion as to the identity of Machilus I mas Rumph., 
Herb. Amb. 3: 68, t. 40, is Teysmann's opinion, quoted by Hass- 



LAURACEAE 



237 



karl, Neue Schlussel (1866) 51, that it belongs in the Lauraceae, 
and of Machilus II femina Rumph. that it might be a species of 
Haasia, which is certainly a wrong disposition of it. 

NEOLITSEA Merrill 
N EOLITSEA AM BOI N ENSIS sp. nov. 

Machilus angustifolia Rumph. Herb. Amb. 7: 60, t. 27, f. 1. 
Amboina, between Soja and Hatalai, Robinson PL Rumph. Amb. 606 
(type), October 24, 1913, in light forests, altitude about 350 meters. 

Arbor circiter 9 m alta, ramulis junioribus inflorescentiisque 
pubescentibus ; foliis verticillatis vel subverticillatis, glabris, 
oblongo-lanceolatis ad oblongo-ellipticis, usque ad 12 cm longis, 
utrinque subaequaliter angustatis, basi acutis, vix triplinervis, 
apice tenuiter acuminatis, nervis utrinque circiter 5, curvato- 
adscendentibus, subtus prominentibus ; fructibus junioribus 
ovoideis, in siccitate nigris, nitidis, circiter 6 mm diametro. 

A tree about 9 m high, the young branchlets and inflorescences 
pubescent. Branches slender, terete, grayish, the younger ones 
reddish-brown, smooth, glabrous, tips of the branchlets rather 
densely appressed-pubescent. Leaves coriaceous, subverticillate 
or verticillate at the ends of the branchlets, oblong-lanceolate 
to oblong-elliptic, 7 to 12 cm long, 2.5 to 3.5 cm wide, subequally 
narrowed to the acute base and to the rather slenderly and 
sharply acuminate apex, the upper surface subolivaceous, smooth 
and shining, the lower slightly paler, sometimes somewhat 
glaucous, glabrous, or the midrib sometimes sparingly pubescent, 
the base scarcely triplinerved ; lateral nerves about 5 on each 
side of the midrib, curved-ascending, scarcely anastomosing, 
prominent on the lower surface, the ultimate reticulations close, 
rather indistinct; petioles more or less pubescent, 5 to 7 mm 
long. Flowers not seen. Infructescences in the axils of fallen 
leaves, fascicled, the pedicels and persistent calyx ferruginous- 
villous, the pedicels rather stout, about 5 mm long. Young fruit 
ovoid, about 6 mm in diameter, glabrous, black and shining 
when dry. 

A species in the group with Neolitsea triplinervia (Blume) 
(Litsea triplinervia Blume), and Neolitsea cassiaefolia (Blume) 
Litsea cassiaefolia Blume), but the leaves not prominently tri- 
plinerved as in these two Javan species, in fact they are strictly 
penninerved, the lower pair being no longer than the next pair 
above. 

The species is, with very little doubt, Machilus angustifolia 
Rumph. Rumphius's description applies closely, but the figure 



238 



RUMPHIUS'S HERBARIUM AMBOINENSE 



is poor and presents the leaves relatively much narrower than in 
the species as here interpreted. Henschel and Pritzel erro- 
neously reduced Machilus angustifolia Rumph. to Tetranthera 
angustifolia WsL\\.==Actinodaphne angustifolia Nees. 

ACTI NODAPH N E Nees 

ACTI NODAPH N E RUM PHI I Blume Mus. Bot. 1 (1851) 344. 

Arbor spiculorum aeruginea Rumph. Herb. Amb. 3: 167, t. 106. 

This species is not represented in our Amboina collections. 
Blume made this reduction of the Rumphian illustration in the 
original description of Actinodaphne rumphii Blume, which was 
based on material from the Moluccas, probably Amboina. He 
cites the Rumphian name as Arbor spiculorum angustifolia, but 
the figure manifestly goes with the form called by Rumphius 
Arbor spiculorum aeruginea. 

ACTINODAPHNE MOLUCCANA Blume Mus. Bot. 1 (1851) 344 (type!). 
Arbor spiculorum latifolia Rumph. Herb. Amb. 5: 167. 

A species of very doubtful status, based wholly on Rumphius's 
description. The third form described by Rumphius in this 
chapter as Arbor spiculorum angustifolia brevifolia is entirely 
doubtful, but is probably a lauraceous plant. 

CRYPTOCARYA R. Brown 

CRYPTOCARYA sp.? 

Lauraster amboinensis maxima Rumph. Herb. Amb. 2: 70, t. 15. 

Nothing resembling this is presented in our Amboina collec- 
tions. The species described and figured by Rumphius is cer- 
tainly a lauraceous plant and is probably a Cryptocarya, judging 
from the rather crude figure. Local names cited by Rumphius 
are leytun, hiber, hiyr, ittir, ayhoo-ittil, and ley itir, so that 
it is very probable that the species can later be located through 
one of the above names. The form described in the same 
chapter as Lauraster amboinensis minor probably represents an 
entirely different species, possibly also a Cryptocarya; its status 
is quite uncertain and cannot be determined from data at present 
available. 

MASSOIA Beccari 

MASSOIA AROMATICA Becc. in d'Albertis New Guinea 2 (1880) 398. 
Cortex oninius s. massoy Rumph. Herb. Amb. 2: 62. 

The reduction has been made from the characters given by 
Rumphius and the name massoy cited by him. The status of 



HERNANDIACEAE 



239 



Massoia as a genus is very uncertain. The form described by 
Rumphius as Cortex oninius II may or may not be the same as 
Massoia aromatica Becc. ; Hasskarl, Neue Schliissel (1866) 28, 
cites it under Cinnamomum kaimis Nees. 

CASSYTHA Linnaeus 

CASSYTHA FILIFORM IS Linn. Sp. PI. (1753) 35. 

Calodium cochinchinensis Lour. Fl. Cochinch. (1790) 247. 

Cussuta s. cussutha indica Rumph. Herb. Amb. 5: 491, t. 184, /• 4- 
Amboina, Hatiwe, Robinson PL Rumph. Amb. 472, September 4, 1913, 
along the seashore. 

This reduction was first made by Linnaeus, in Stickman Herb. 
Amb. (1754) 25, Amoen. Acad. 4 (1759) 133, Syst. ed. 10 (1759) 
862, Sp. PI. ed. 2 (1762) 531, and this is manifestly the correct 
disposition of Cussutha indica Rumph. Calodium cochinchin- 
ensis Lour., under which Loureiro cited the Rumphian descrip- 
tion and illustration, is a synonym of Cassytha filiformis Linn. 

HERNANDIACEAE 

HERNANDIA Plumier 

HERNANDIA OVIGERA Linn, in Stickman Herb. Amb. (1754) 14, Amoen. 
Acad. 4 (1759) 125, Syst. ed. 10 (1759) 1264, Sp. PL ed. 2 (1763) 
1392 (type!). 

Arbor ovigera femina Rumph. Herb. Amb. 3: 193, t. 123. 
Amboina, Robinson PI. Rumph. Amb. 476, a single specimen separated 
from No. 477, Hernandia peltata Meisn. 

This represents the form figured by Rumphius and the one 
described as Arbor ovigera femina. The Linnean species is 
typified by the Rumphian figure and description, Hernandia 
ovigera Linn, being based wholly on Arbor ovigera t. 123. 

HERNANDIA PELTATA Meisn. in DC. Prodr. 15 1 (1864) 263. 
Arbor ovigera mas Rumph. Herb. Amb. 3: 193. 
Amboina, Robinson PI. Rumph. Amb. 477, August 8, 1913, along the 
seashore near the town of Amboina, locally known as mata ipang or mata 
ikang. 

The form described by Rumphius as Arbor ovigera mas is un- 
mistakably Hernandia peltata Meisn., the leaves being definitely 
described as peltate. It is strongly suspected that Hernandia 
peltata Meisn. is not specifically distinct from Hernandia ovigera 
Linn., but is merely a form with peltate leaves. It is assumed 
that the specimen cited under Hernandia ovigera Linn, and the 
one cited under H. peltata Meisn. came from the same tree. 



240 RUMPHIUS'S HERBARIUM AMBOINENSE 

CRUCIFERAE 

NASTURTIUM * R. Brown 

NASTURTIUM INDICUM (Linn.) DC. Syst. 2 (1821) 199. 

Sisymbrium indicum Linn. Mant. 1 (1767) 93. 

Sinapi indigenum s. amboinicum Rumph. Herb. Amb. 5: 282. 

Amboina, near the town of Amboina, Robinson PL Rumph. Amb. 268, 
July 19, 1913, in and near ditches. 

Hasskarl, Neue Schlussel (1866) 121, has suggested that this 
is a species of Sinapis; but from Rumphius's brief description, 
the indicated habitat, and the note that it has no economic use, 
it is evident that Nasturtium is the plant intended. 

BRASSICA (Tourn.) Linnaeus 

BRASSICA JUNCEA (Linn.) Coss. in Bull. Soc. Bot. France 6 (1859) 609. 

Sinapis juncea Linn. Sp. PI. (1753) 668. 
Sinapi sinense Rumph. Herb. Amb. 5: 282. 

Amboina, Titauuku, Robinson PL Rumph. Amb. J+05, October 8, 1913, 
in cleared places, altitude about 120 meters, locally known as susawi 
ambon and rumput china blanda. 

This specimen I take to be typical Brassica juncea Coss., the 
ordinary form that is widely distributed in the Malay Archipel- 
ago. The plant, or plants, actually described by Rumphius, are, 
however, the cultivated forms of Chinese origin, which appear 
to be forms of Brassica pekinensis (Lour.) Skeels (Sinapis 
pekinensis Lour.). Rumphius indicates two forms, under the 
names album and nigrum. It seems very probable that this 
commonly cultivated Chinese form is but a horticultural variety 
of Brassica juncea Coss. 

CAPPARIDACEAE 

POLANISIA Rafinesque 

POLANISIA VISCOSA (Linn.) DC. Prodr. 1 (1824) 242. 

Cleome viscosa Linn. Sp. PI. (1753) 672. 
Lagansa alba Rumph. Herb. Amb. 5 : 280, t. 96, f. 3. 
Amboina, Liang, Robinson PL Rumph. Amb. U12, November 29, 1913, 
along roadsides, at low altitudes, locally known as lagansa. 

This common weed was originally reduced by Linnaeus, in 



* Retained name, Brussels Congress; Cardaminium Moench (1794), and 
Baeumerta Gaertn. (1800) are older. 



> 



MORINGACEAE 241 

Stickman Herb. Amb. (1754) 21, Amoen. Acad. 4 (1759) 130, 
Syst. ed. 10 (1759) 1137, to Cleome icosandra Linn., but the 
plant described as Lagansa alba is manifestly the form com- 
monly known as Polanisia viscosa (Linn.) DC. In the descrip- 
tion of the plate figure 2 is connected with Lagansa alba, but 
it is apparent that the figures are reversed between Lagansa 
alba and Lagansa rubra. 

GYNANDROPSIS * de Candolle 

GYNANDROPSIS PENTAPHYLLA (Linn.) DC. Prodr. 1 (1824) 238. 

Cleome pentaphylla Linn. Sp. PI. ed. 2 (1763) 938. 
Lagansa rubra Rumph. Herb. Amb. 5: 280, t. 96, f. 2. 

The common and well-known Gynandropsis pentaphylla DC. 
is not represented in our Amboina collections. The Rumphian 
description, however, applies well to this species. In the descrip- 
tion of the plate the explanations of Lagansa alba and Lagansa 
rubra are transposed. 

MORINGACEAE 

MORINGA Burman f. 

MORINGA OLEIFERA Lam. Encycl. 1 (1785) 398. 

Guilandina moringa Linn. Sp. PI. (1753) 381. 
Moringa pterygosperma Gaertn. Fruct. 2 (1791) 314. 
Morunga Rumph. Herb. Amb. 1: 184, t. 7-4. 
Morunga femina Rumph. Herb. Amb. 1: 185, t. 75. 

Amboina, in cultivation, near the town of Amboina, Robinson PI. Rumph. 
Amb. 255, September 13, 1913, locally known as kelor. 

The reduction of Morunga, t. 7U, was first made by Linnaeus, 
in Stickman Herb. Amb. (1754) 8, Amoen. Acad. 4 (1759) 
120; and in the Systema, ed. 10 (1759) 1018, he also reduced 
Morunga femina, t. 75. Both figures manifestly represent the 
same species, and both are unmistakably the common and well- 
known Moringa oleifera Lam. Other names, all synonyms, to 
which the Rumphian figures have been referred by various au- 
thors are : Hyper anther a moringa Vahl, Moringa. zeylanica Willd., 
Moringa domestica Ham., Anoma moringa Lour., A. morunga 
Lour., and Moringa polygona DC. ; some of these are to be inter- 
preted in part by the Rumphian figures. Moringa domestica 
Ham., in Mem. Wern. Soc. 5 2 (1826) 368, 371, does not appear 
in Index Kewensis. 



* Retained name, Vienna Code; Pedicellaria Schrank (1790) is older. 

144971 16 



242 



RUMPHIUS'S HERBARIUM AMBOINENSE 



NEPENTHACEAE 

NEPENTHES Linnaeus 

NEPENTHES M I RABi LIS (Lour.) Merr. comb. nov. 

Phyllamphora mirabilis Lour. Fl. Cochinch. (1790) 606. 

Nepenthes phyllamphora Willd. Sp. PI. 4 2 (1805) 874. 

Cantharifera Rumph. Herb. Amb. 5: 121, t. 59, f. 2. 
Amboina, Batoe mera, Robinson PL Rumph. Amb. 256, July 31, 1913, on 
a fern-covered hillside along the river at an altitude of from 10 to 50 
meters; Batoe gadjah, Robinson PI. Rumph. Amb. 257, August 5, 1913, on 
grassy hillsides at an altitude of about 150 meters. 

Cantharifera was erroneously reduced by Linnaeus to Nepen- 
thes distillatoria Linn., the type of the genus, and a species 
confined to Ceylon, in Stickman Herb. Amb. (1754) 20, Amoen. 
Acad. 4 (1759) 129, Syst. ed. 10 (1759) 1247, Sp. PI. ed. 2 
(1763) 1354, in which he was followed by Burman f., Fl. Ind. 
(1768) 190. Loureiro, however, Fl. Cochinch. (1790) 606, 
described a Cochin-China specimen as Phyllamphora mirabilis 
Lour., and under it he discussed Cantharifera Rumph., stating 
that it differed from his plant in its prostrate stems and scandent 
branches. The species, as described by Loureiro, must be inter- 
preted from Cochin-China specimens, but in all probability is 
the same as the widely distributed species commonly known as 
Nepenthes phyllamphora Willd., which is definitely known from 
low altitudes from southern China to the Philippines, Borneo, 
Malay Peninsula, Sumatra, Amboina, the Caroline Islands, and 
New Guinea. Nepenthes phyllamphora Willd. was based wholly 
on Loureiro's species, but Willdenow definitely refers here Can- 
tharifera Rumph. Under the accepted code of botanical nomen- 
clature the above new combination is necessary for this 
well-known and widely distributed species ; both names, Phyllam- 
phora mirabilis and Nepenthes phyllamphora, must be inter- 
preted by the plant Loureiro described. 

NEPENTHES MAXIMA Reinw. ex Nees in Ann. Sci. Nat. I 3 (1824) 369, 
t. 20, f. 3. 

Cantharifera alba Rumph. Herb. Amb. 5: 122. 
Amboina, Salahoetoe, Robinson PI. Rumph. Amb. 413, November 27, 1913, 
terrestrial and climbing, chiefly at an altitude of about 900 meters, locally 
known as tampayan utan. 

The identification of Cantharifera alba with Nepenthes maxi- 
ma Reinw. was suggested by Hasskarl, Neue Schlussel (1866) 
103, and there is but very little doubt that this is the correct 
disposition of it. It was described by Rumphius from specimens 
received from the neighboring island of Little Ceram and was 



CRASSULACEAE— PITTQSPORACEAE 



243 



not recorded by him as being from Amboina. Macfarlane * 
definitely records Nepenthes maxima Reinw. from Amboina, 
New Guinea, Celebes, and Borneo. 

CRASSULACEAE 

KALANCHOE Adanson 

KALANCHOE LACINIATA (Linn.) DC. PI. Grass. (1799-1829) t. 100, 
Prodr. 3 (1828) 395. 
Cotyledon laciniata Linn. Sp. PI. (1753) 430. 
Planta anatis Rumph. Herb. Amb. 5: 275, t. 95. 

This species is not represented in our Amboina collections. 
The figure is poor, but the description, at least in most part, 
applies to this widely distributed species. It may, in part, 
apply also to Bryophyllum pinnatum (Lam.) Kurz (B. calycinum 
Salisb.) , but the description of the flowers as 5-merous and yellow 
indicates a Kalanchoe not a Bryophyllum. The reduction of 
Planta anatis to Cotyledon laciniata Linn, was made by Lin- 
naeus himself, in Stickman Herb. Amb. (1754) 21, Amoen. 
Acad. 4 (1759) 130, Syst. ed. 10 (1759) 1036, Sp. PL ed. 2 
(1762) 615, in which he has been followed by all later authors, 
either under Cotyledon, Kalanchoe, or Verrea. 

Bryophyllum pinnatum (Lam.) Kurz. (B. calycinum Salisb.) 
occurs in our Amboina collections (Roemah tiga, Rel. Robins. 
1819, July 30, 1913) and, as noted above, may be included in the 
Rumphian description of Planta anatis. Primarily, however, 
Planta anatis is certainly Kalanchoe laciniata (Linn.) DC. 

PITTOSPORACEAE 

PITTOSPORUM Banks 
PITTOSPORUM MOLUCCANUM (Lam.) Miq. 111. Fl. Arch. Ind. (1871) 76. 
Anasser moluccana Lam. 111. 2 (1797) 40 (type!). 
Anassera moluccana Pers. Syn. 1 (1805) 265 (type!). 
Pittosporum ferrugineum Ait. var. filarium DC. Prodr. 1 (1824) 847 
(type!). 

Pittosporum rumphii Putterl. Syn. Pitt. (1839) 7 (type). 
Cortex foetid us Rumph. Herb. Amb. 7: 12, t. 7. 

This species is not represented in our Amboina collections, 
but Miquel, 1. c, cites Amboina material collected by Zippel 
and by Teysmann, on which his sufficiently ample description is 
based. The Rumphian figure and description are the whole 
basis of Anasser moluccana Lam., Anassera moluccana Pers., 
and de Candolle's variety of Pittosporum ferrugineum cited 
above (erroneously cited as Cortex filiarius, but the page and 



* Engl. Pflanzenreich 36 (1908) 76. 



244 



RUMPHIUS'S HERBARIUM AMBOINENSE 



plate numbers are correct), and apparently Putterlick's Pit- 
tosporum rumphii. It differs from Pittosporum ferrugineum 
Ait. notably in its larger fruits and leaves. 

PITTOSPORUM sp.? 

Cortex igneus Rumph. Herb. Amb. 7: 10, t. 6, f. 1. 

Hasskarl, Neue Schliissel (1866) 186, thought that this might 
be a species of Pittosporum, following Teysmann's suggestion. 
It was not from Amboina, but from the Aru Islands or from 
others in the same general region. Native names cited are culit 
apt, eyk, and mockulhaulo. There is little in the description to 
indicate that it is a Pittosporum, and its status is quite unde- 
terminable from the data and the material at present available. 

CUNONIACEAE 

WEINMANNIA * Linnaeus 

WEINMANNIA FRAXINEA Sm. ex D. Don in Edinb. New Philos. Journ. 9 
(1830) 93. 
Pterophylla fraxinea D. Don 1. c. 
Cortex papetarius Rumph. Herb. Amb. 3: 212, t. 137. 

The type of Weinmannia fraxinea Sm. was from the Moluccas, 
so that the probabilities are that Teysmann was correct in his 
reduction of Cortex papetarius Rumph. to Smith's species. 
Rumphius's figure is an excellent one and is unmistakably a 
Weinmannia. Linnaeus, Mant. 2 (1771) 510, erroneously 
reduced it to Dialum indum Linn., in which he was followed by 
numerous authors. Burman f., Fl. Ind. (1768) 12, considered 
that it represented a variety of Dialum javanicum Burm. f., 
which is a synonym of D. indum Linn. Hasskarl, Neue Schliissel 
(1866) 68, thought that it might be an Otonychium=Harpullia 
(Sapindaceae) , with which it has nothing in common, and at 
the same time quotes Teysmann's opinion that it is Weinmannia 
fraxinea Sm. 

SCHIZOMERIA D. Don 

SCHI20MERIA SERRATA Hochr. in Ann. Conserv. Jard. Bot. Geneve 10 
(1907) 118. 

Acronychia serrata Hochr. PI. Bogor. Exsicc. (1904) 49. 
Tanarius major Rumph. Herb. Amb. 3: 192, t. 122. 
Arbor vespertilionum II oppositifolia Rumph. Herb. Amb. 7: 17, t. 10. 
Amboina, Robinson PI. Rumph. Amb. 60S, without definite locality or date. 

The specimen is a good match for the figure of Tanarius major 
Rumph. and agrees with the description even better than with 
the figure. Tanarius major Rumph. was discussed by Loureiro, 



* Retained name, Vienna Code; Windmannia P. Br. (1756) is older. 



HAMAMELIDACEAE— ROSACEAE 



245 



Fl. Cochinch. (1790) 231, under Jambolifera resinosa Lour., 
and has been cited by Don, Henschel, and de Candolle under 
Cyminosma resinosa Don, but has nothing to do with the species 
that Loureiro described. The specimen also agrees perfectly 
with the figure of Arbor vespertilionum Rumph. and with the 
short description of the second plant included in the descrip- 
tion. The major part of the description of Arbor vespertilio- 
num Rumph., that is, the first form described in this chapter, 
is Helicia serrata R. Br. (see p. 205), and the figure has 
been referred to Helicia, where it manifestly does not belong. 
Schizomeria serrata Hochr. is known only from Amboina, was 
originally described from specimens cultivated in the botanic 
garden at Buitenzorg, Java, and is very closely allied to the 
Australian species Schizomeria ovata D. Don, the only other 
known species of the genus. This is one of the few cases in 
which Rumphius described and figured the same species twice 
under entirely different names. 

HAMAMELIDACEAE 

ALTINGIA Noronha 

ALTINGIA EXCELSA Noronha in Verh. Batav. Genootsch. 5 a (1795) 9. 
Lignum papuanum I Rumph. Herb. Amb. 2: 57? 

Persoon, Syn. 2 (1807) 579, reduced this to Altingia excelsa 
Noronha, apparently after Noronha, and Blume reduced it to 
Liquidamber altingia B1.=L. altingiana Blume, both synonyms of 
Altingia excelsa Noronha. It is probable that Rumphius in- 
cluded more than this one species under Lignum papuanum 7, 
especially in view of the fact that Altingia excelsa Noronha does 
not appear to be definitely known from so far to the east as 
New Guinea. The form described in the same chapter as 
Lignum papuanum II, unless referable to Altingia excelsa Noronha, 
is undeterminable. The identification of Lignum papuanum 
is based largely on the native name cited by Rumphius, caju 
rasamala; in Java rasamala is the resin of Altingia excelsa 
Noronha. The plant Rumphius had in mind may have been 
entirely different. 

ROSACEAE 

RUBUS Linnaeus f 

RUBUS MOLUCCANUS Linn. Sp. PL (1753) 1197 (type!). 

Rubus moluccus latifolius Rumph. Herb. Amb. 5: 88, t. 47, /. 2. 
Amboina, Halong and near the town of Amboina, Robinson PL Rumph. 
Amb. 270, July 16 and 23, 1913, in light forests and open places, altitude 
about 10 meters, locally known as buan tampayang and daun doeri doeri. 



246 



RUMPHIUS'S HERBARIUM AMBOINENSE 



The Rumphian figure and description are the whole basis of 
Rubus moluccanus Linn., but this species has been interpreted 
by many authors as a polymorphous one and given a range from 
the Himalayan region to Ceylon, southern China, the Philippines, 
Malaya, and northeastern Australia. Hooker f., Fl. Brit. Ind. 
2 (1878) 330, states: 

I am quite unable to arrange the forms of this common and protean 
plant under recognizable varieties answering to its synonymy. 

Focke, in his monograph of the genus Rubus, Bibl. Bot. 17 1 
(1910) 89, reproduces the Rumphian figure and limits the 
species to Amboina. He cites no specimens, gives a short de- 
scription after data given by Rumphius, and was unable to 
determine the status of the species in a satisfactory manner. 
He states: 

Die weitere Verbreitung ist vollig unsicher, weil die Art zu ungeniigend 
bekannt ist. Anscheinend gehoren hieher Exemplare von Voun auf Neu- 
guinea (leg. Tejsmann) und vielleicht auch von Luzon. 

Rubus moluccanus Linn, has been treated by many authors as 
a collective species, and it is very evident that numerous forms 
so named in herbaria cannot be properly placed under this 
species, but must be considered as distinct ones. The Amboina 
specimens closely match a large series of specimens in the 
herbarium of the Bureau of Science from various parts of the 
Philippines and a few specimens from Borneo and Java. It 
is suspected that the typical form will be found to be of wide 
distribution in the Malay Archipelago. Its exact status can now 
be definitely determined by a direct comparison with the topotype 
cited above. All authors have followed Linnaeus in the reduc- 
tion of the Rumphian figure and description, so that the case 
is not complicated by synonyms, except as species described by 
later authors, without reference to Rumphius, may have been 
reduced to typical Rubus moluccanus Linn. 

RUBUS FRAXINIFOLIUS Poir. in Lam. Encycl. 6 (1804) 242, subsp. 
CELEBICUS (Blume) Focke in Bibl. Bot. 17 2 (1911) 150, /. 63. 

Rubus celebicus Blume Bijdr. (1826) 1107. 

Rubus parvifolius Linn. Sp. PI. (1753) 1197 p. p. (quoad syn. Rumph.). 
Rubus moluccus parvifolius Rumph. Herb. Amb. 5: 88, t. 1>7, f. 1. 
Amboina, near the town of Amboina, Robinson PI. Rumph. Amb. 271, 
July 23, 1913, hillsides and river banks, altitude 5 to 10 meters. 

The Rumphian figure and description were cited by Linnaeus 
in the original description of Rubus parvifolius, Sp. PL (1753) 



ROSACEAE 



247 



1197, but the actual type, on which the description was based 
and from which the Linnean species must be interpreted, was 
a specimen collected near Canton, China, by Osbeck. This speci- 
men is the same as Rubus triphyllus Thunb., Fl. Jap. (1784) 
215, the name that Focke has adopted for the species. However, 
Focke is manifestly in error in the selection of this name as 
the valid one for the species is Rubus parvifolius Linn. (1753) 
[not R. parviflorus Linn, as cited by Focke, Bibl. Bot. 17 2 (1911) 
187]. Rubus fraxinifolius Poir. was described from Javan 
specimens collected by Commerson and is widely distributed in 
the Sunda Islands, especially in Java and Sumatra. The sub- 
species celebicus Focke is widely distributed in the Philippines 
and in the Moluccas, extending to New Guinea. 

PARINARIUM Aublet 

PAR I N AR I U M GLABERRI MUM Hassk. in Tijdschr. Nat. Ges. 10 (1843) 
147, nomen nudum, Cat. Hort. Bogor. (1844) 269, nomen nudum, 
Flora 27 (1844) 583. 
Parinarium scabrum Hassk. in Tijdschr. Nat. Ges. 10 (1843) 147, 
nomen nudum, Cat. Hort. Bogor. (1844) 269, nomen nudum, Flora 
27 (1844) 585. 

Parinarium laurinum A. Gray Bot. Wilkes U. S. Explor. Exped. (1854) 
490, t. 55. 

Parinarium ellipticum T. & B. Cat. Hort. Bogor. (1866) 253. 
Parinarium macrophyllum T. & B. 1. c, Nat. Tijdschr. Ned. Ind. 29 
(1867) 256. 

Parinarium wAndanaense Perk. Frag. Fl. Filip. (1904) 119. 
Parinarium racemosum Merr. in Govt. Lab. Publ. (Philip.) 17 (1904) 
19, non Vid. 

Parinarium curranii Merr. in Philip. Journ. Sci. 4 (1910) > Bot. 264. 

Atunus Rumph. Herb. Amb. 1: 171, t. 66. 
Amboina, Way tommo, Robinson PI. Rumph. Amb. 273, August 16, 
1913, in open forests, altitude about 5 meters, in fruit; Amboina (town), 
Robinson PI. Rumph. Amb. 272, November 4, 1913, from a cultivated tree, 
in flower. Both specimens bear the common name atun. 

Atunus of Rumphius has not been previously referred to its 
proper place in our present system of classification, although 
Hasskarl, Neue Schlussel (1866) 22, suggested that it was a 
Parinarium. It manifestly is Parinarium, and the species very 
generally known as Parinarium scabrum Hassk., of which 
numerous synonyms are given above. In citing the above syn- 
onyms I have in part followed Koorders and Valeton, Bijdr. 
Boomsoort. Java 5 (1900) 337, but have also seen authentically 
named specimens of most of the species I have here reduced. 
Loureiro,. Fl. Cochinch. (1790) 295, mentions it, following the 



248 



RUMPHIUS'S HERBARIUM AMBOINENSE 



description of Stixis scandens Lour., as apparently belonging in 
the same genus as that species, but Atunus has nothing in 
common with Stixis scandens Lour., which belongs in the Cap- 
paridaceae. Atunus albus Rumph., 1. c. 172, is probably a form 

of P armarium glaberrimum Hassk. 

CONNARACEAE 

CONNARUS Linnaeus 

CONNARUS sp. 

Clompanus funicularis Rumph. Herb. Amb. 5: 70. t. 37, f. 2. 

The description and figure are certainly those of a Connarus, 
and perhaps Clompanus funicularis Rumph. is the same as Con- 
narus gaudichaudii Planch. Lamarck, Encycl. 2 (1786) 52, 
places it under Clompanus paniculatus Aubl., a species based 
on material from tropical America, and one that has nothing to 
do with the form that Rumphius figured and described. Miquel, 
Fl. Ind. Bat. I 1 (1855) 349, suggested that it might be a species 
of Millettia, where it certainly does not belong. 

LEGUMINOSAE 

PITHECOLOBIUM * Martius 

PITHECOLOBIUM CLYPEARIA (Jack) Benth. in Hook. Lond. Journ. Bot. 3 
(1844) 209. 

Jnga clypearia Jack Malay Miscel. 2 (1822) 78. 

Mimosa trapezifolia Roxb. Hort. Beng. (1814) 93, nomen nudum, Fl. 

Ind. ed. 2, 2 (1832) 546. 
Adenanthera circinalis DC. Prodr. 2 (1825) 446 (type!). 
Clypearia rubra Rumph. Herb. Amb. 3: 176, t. 112. 
Amboina, Koesoekoesoe sereh, Robinson PL Rumph. Amb. 548, October 
3, 1913, in light forests, altitude about 285 meters. 

In proposing the name Adenanthera falcata Linn., in Stickman 
Herb. Amb. (1754) 14, Amoen. Acad. 4 (1759) 124, Linnaeus 
referred to it both Clypearia alba and Clypearia rubra as rep- 
resented by tt. Ill and 112 of Rumphius. I maintain that 
the name must be typified by the first figure mentioned, that is 
t. Ill, especially in view of the fact that in the Syst. ed. 10 
(1759) 1020 and Sp. PL ed. 2 (1762) 550 Linnaeus excluded 
under Adenanthera falcata and A. falcataria the reference to 
t. 112, Clypearia rubra, limiting the species to Clypearia alba 
Rumph. t. 111. I have not seen the original description of 
Inga clypearia Jack, which was based on Sumatran specimens; 



* Retained name, Brussels Congress; Zygia Boehm. (1760) is older. 



LEGUMINOSAE 



249 



the specific name, however, was unquestionably taken from 
Rumphius. Prain * cites Inga clypearia Jack as a doubtful 
synonym of Pithecolobium clypearia Benth., but if Inga clypearia 
Jack should prove to be different from Pithecolobium clypearia 
Benth. as currently interpreted, then the specific name will go 
with Jack's species. Adenanthera circinnalis DC. is based 
wholly on Clypearia rubra Rumph., and both are cited by 
Bentham as synonyms of Pithecolobium clypearia (Jack) Benth., 
in Trans. Linn. Soc. 30 (1875) 580. 

ALBIZZIA Durazzini 

ALBIZZIA SAPONARIA (Lour.) Blume ex Miq. Fl. Ind. Bat 1 1 (1855) 19. 

Mimosa saponaria Lour. FL Cochinch. (1790) 653. 

Inga saponaria Willd. Sp. PI. 4 (1805) 1008. 

Cortex saponarius Rumph. Herb. Amb. 4: 131, t. 66. 
Amboina, Amahoesoe, Robinson PL Rumph. Amb. 523, August 13, 1913, 
on limestone formation, altitude about 30 meters, in flower; Hatiwe, 
Robinson PL Rumph. Amb. 52U, September 15, 1913, in forests, altitude 
about 300 meters, in fruit, locally known as langir. 

Mimosa saponaria Lour, was described from a Cochin-China 
specimen, which has been universally considered, and probably 
is, identical with the form that Rumphius described. Loureiro 
also cites the Rumphian plant as representing his species, while 
Cochin-China material in various herbaria is identical with the 
common Philippine and Moluccan form of the species. 

ALBIZZIA FALCATA (Linn.) Backer comb. nov. 

Adenanthera falcata Linn, in Stickman Herb. Amb. (1754) 14, Amoen. 

Acad. 4 (1759) 124, Syst. ed. 10 (1759) 1020 (type!). 
Adenanthera falcataria Linn. Sp. PI. ed. 2 (1762) 550 (type!). 
Albizzia moluccana Miq. Fl. Ind. Bat. 1 1 (1855) 26; Koord. in Meded. 

Lands Plantent. 19 (1898) 419. 
Clypearia alba Rumph. Herb. Amb. 3: 176, t. 111. 

This species is not represented in our Amboina collections. 
However, after a very careful study of Rumphius's description 
I have definitely concluded that the form he described and 
figured is identical with Albizzia moluccana Miq., a species 
originally described from leaf specimens only, but of which 
Koorders has given an ample and detailed description. Mr. 
Backer in connection with his study of the Leguminosae of 
Java has independently reached the same conclusion. In all 
botanical literature the Rumphian Clypearia alba has been cited 
under Adenanthera falcata Linn, and A. falcataria Linn., both 

*Journ. As. Soc. Beng. 66 2 (1897) 274. 



4 



250 RUMPHIUS'S HERBARIUM AMBOINENSE 

being based wholly on Rumphius's description and figure. The 
description, compiled wholly from Rumphius, is given under one 
or the other of these names by Burman f., Fl. Ind. (1768) 101, 
Murr. Syst. (1771) 398, Lam. Encycl. 2 (1786) 76, Pers. Syn. 
1 (1805) 461, Willd. Sp. PL 2 (1799) 550, DC. Prodr. 2 (1825) 
446, Don Gen. Syst. 2 (1832) 399, Dietr. Syn. 2 (1839-52) 1425, 
Walp. Repert. 5 (1846) 580, Miq. Fl. Ind. Bat. I 1 (1855) 47, 
and Walp. Ann. 4 (1857) 613, yet the status of the species has 
been very doubtful up to the present time. 

ALBIZZIA RETUSA Benth. in Hook. Lond. Journ. Bot. 3 (1844) 90. 

Albizzia littoralis T. & B. in Nat. Tijdschr. Ned. Ind. 29 (1866) 259. 
Clypearia maritima Rumph. Herb. Amb. 3: 199. 

This species is not represented in our Amboina collections. 
I have very little doubt that this is the correct disposition of 
Clypearia maritima, after the description and data given by 
Rumphius. Hasskarl, Neue Schlussel (1866) 66, follows Rader- 
macher in considering it an undetermined species of Adenan- 
thera. The species is scattered along the seashore from the 
Nicobar Islands through Malaya and the Philippines to the 
Caroline Islands and has been reported from Amboina. 

ALBIZZIA PROCERA (Roxb.) Benth. in Hook. Lond. Journ. Bot. 3 (1844) 

89. 

Mimosa procera Roxb. PI. Coromandel 2 (1798) 12, t. 121. 
Lignum murinum majus Rumph. Herb. Amb. 3: 50, t. 28. 

Nothing resembling this occurs in our Amboina collections, 
but I have not the slightest doubt that the plant figured is iden- 
tical with Albizzia procera Benth., which extends from India 
to tropical Australia. Miquel, Fl. Ind. Bat. I 1 (1855) 54, has 
suggested that the Rumphian figure represents an Albizzia, but 
otherwise no other author has suggested an identification of 
Lignum murinum. 

The two other forms described, but not figured by Rumphius, 
1. c. 50, 51, as Lignum murinum minus and Lignum murinum 
parvifolium, probably also represent species of Albizzia, as sug- 
gested by Miquel. Their status, however, cannot be definitely 
determined from data at present available. 

Caju ticcos leytimorensis Rumph., Herb. Amb. 3 : 52, briefly de- 
scribed by Rumphius, is suggested by Hasskarl, Neue Schlussel 
(1866) 50, as a possible synonym of Albizzia procera Benth. 
The description of the pods and of the wood hardly conforms to 
Roxburgh's species ; it probably is an Adbizzia, but its status must 
remain doutbful pending further exploration of Amboina. 



LEGUMINOSAE 



251 



ALBIZZIA sp. 

Clypearia rubra s. sye II Rumph. Herb. Amb. 3: 177. 

The description is very brief, following that of Clypearia 
rubra. Hasskarl, Neue Schliissel (1866) 205, suggests that it 
may be Albizzia moluccana Miq., which I have here reduced to 
Albizzia falcata (Linn.) Backer. Its exact status is indetermin- 
able at the present time, but it is probably a species of Albizzia. 

ACACIA Linnaeus 

ACACIA MANGIUM Willd. Sp. PL 4 2 (1805) 1053 (type!). 

Mimosa simplici folia Linn. var. mangium Poir. in Lam. Encycl. Suppl. 
1 (1810) 61. 

Acacia holosericea A. Cunn. ex G. Don Gen. Syst. 2 (1832) 407. 
Mangium montanum Rumph. Herb. Amb. 3: 123, t. 81. 

A species based wholly on Rumphius's description and figure. 
The description, after Rumphius, has been repeated by de 
Candolle, Prodromus 2 (1825) 451; Don, Gen. Syst. 2 (1832) 
403; Miquel, Fl. Ind. Bat. I 1 (1855) 15; and Bentham, Trans. 
Linn. Soc. 30 (1875) 495. Bentham thought that it was probably 
allied to the Australian Acacia holosericea A. Cunn., and I con- 
sider his surmise correct; in fact I can detect no differences 
between Australian specimens and material from the Island of 
Buru (cult. Buitenzorg I-C-37-K-32) . Forster f., Prodr. 
(1786) 75, referred Mangium montanum Rumph. to Mimosa 
mangium Forst. f., basing his description, however, on actual 
specimens from the Friendly Islands, New Caledonia, and New 
Hebrides. Acacia mangium Willd. was published independently 
of the earlier Mimosa mangium Forst. f . 

ACACIA RUGATA (Lam.) Ham. in Wall. Cat. (1832) no. 5251. 
Mimosa rug ata Lam. Encycl. 1 (1783) 20. 
Mimosa concinna Willd. Sp. PI. 4 (1805) 1039. 
Acacia concinna DC. Prodr. 2 (1825) 464. 
Guilandina microphylla DC. Prodr. 2 (1825) 480 (type!). 
Nugae silvarum minimae Rumph. Herb. Amb. 5: 95, t. 49, f. 2. 

This species is not represented in our Amboina collections. 
The Rumphian figure is poor, presenting a sterile branch only, 
but both it and the description conform better to Acacia rugata 
(Lam.) Ham. than to any other known species, so that the 
present reduction is probably the correct disposition of it. Bur- 
man f., Fl. Ind. (1768) 99, thought it represented a variety of 
Guilandina nuga Linn. Guilandina microphylla DC. was based 
wholly on the Rumphian figure and description. Wight and 
Arnott, Prodr. (1834) 277, reduced it to Acacia concinna DC. 



252 



RUMPHIUS'S HERBARIUM AMBOINENSE 



with doubt, while Miquel cites it as a possible synonym of 
Acacia hooperiana Zipp. var. subcuneata (Bl.) Miq. Ft. Ind. 
Bat. T (1855) ll=Acacia concinna DC. =Acacia rugata (Lam.) 
Ham. 

M I M OSA Linnaeus 

MIMOSA PUDICA Linn. Sp. PI. (1753) 518. 

Herba mimosa Rumph. Herb. Amb. 5: 303. 
Amboina, Robinson PL Rumph. Amb. 526, July 25, 1913, along small 
streams near the town of Amboina. 

The identification of Herba mimosa follows Hasskarl, Neue 
Schliissel (1866) 124, which is certainly the correct disposition 
of it. This weed originated in tropical America and is now 
found in most hot countries. As noted by Doctor Robinson this 
was known to Rumphius only by description, but is now fairly 
common in Amboina. 

In the discussion of the various species of plants with sensitive 
leaves, following Herba sentiens Rumph., Herb. Amb. 5 : 301, 
several representatives of the Mimosoideae are briefly described 
or mentioned, some probably are species of the genus Mimosa, 
others may belong in allied genera. Following Hasskarl, 
Neue Schliissel (1866) 124, 125, Caban cabanan, page 304, may be a 
Mimosa; Similis planta peruana, page 304, may be Mimosa dor miens 
HBK. ; Altera planta peruana, page 304, may be Mim^osa humilis 
HBK. ; Pina hui huitzli, page 304, may be Mimosa casta Linn.; 
Arbor pudica, page 305, may be Mimosa pudibunda Willd. ; while 
Planta sentiens hispanorum and Herba viva are wholly doubtful. It 
is hardly worth while to consider these forms, as the data given 
are in most cases quite insufficient on which to base a definite 
identification of the several forms; and Hasskarl's determina- 
tions, as any must be, are for the most part merely guesses. 
They were not from Amboina and were known to Rumphius by 
description only. 

ADENANTHERA Royen 

ADENANTHERA PAVONINA Linn. Sp. PL (1753) 384. 

Corallaria parvifolia Rumph. Herb. Amb. 3: 173, t. 109. 

Amboina, Paso, Robinson PI. Rumph. Amb. 5^, October 29, 1913, along 
the seashore. 

The original reduction of Corallaria parvifolia to Adenanthera 
pavonina was made by Linnaeus, in Stickman Herb. Amb. (1754) 
14, is the correct disposition of it, and has been consistently 
followed by all subsequent authors who have had occasion to cite 
the Rumphian figure. 



LEGUMINOSAE 



253 



ENTADA * Adanson 

ENTADA PHASEOLOIDES (Linn.) Merr. in Philip. Journ. Sci. 9 (1914) 
Bot. 86. 

Lens phaseoloides Linn, in Stickman Herb. Amb. (1754) 18, Amoen. 
Acad. 4 (1759) 128 (type!). 

Mimosa entada Linn. Sp. PI. (1753) 518. 

Mimosa scandens Linn. Sp. PI. ed. 2 (1763) 1501. 

Entada scandens Benth. in Hook. Lond. Journ. Bot. 4 (1842) 332. 

Entada rumphii Scheff. in Nat. Tijdschr. Nederl. Ind. 32 (1871) 412. 

Faba marina major Rumph. Herb. Amb. 5: 5, t. J*. 
Amboina, Amahoesoe, Robinson PL Rumph. Amb. 528, August, 28, 1913, 
in flower, growing near the beach; Kati-kati, Robinson PI. Rumph. Amb. 
529, October 7, 1913, in flower and fruit, growing at an altitude of about 
80 meters. 

The Rumphian plate is the whole basis of Lens phaseoloides 
Linn., which supplies the oldest valid specific name for this 
widely distributed species, although the name Lens phaseoloides 
as published by Linnaeus was probably taken from Burman, 
Thesaurus Zeylanicus (1737) 139. The original publication in 
Stickman's Herbarium Amboinense is as follows : "5. Faba 
marina. Lens phaseoloides ; proprii generis." As to the pro- 
priety of taking up the generic name Lens in place of Entada, as 
W. F. Wight proposes,! I have already fully discussed the mat- 
ter and do not believe that Mr. Wight's proposal merits the 
approval of ' botanists. J 

The form described by Rumphius as Parrana nigra, Herb. Amb. 
5 : 7, may be an Entada as suggested by Hasskarl, Neue Schliissel 
(1866) 89, but at any rate it is apparently entirely different 
from Entada phaseoloides Merr. The description is too im- 
perfect to warrant definite determination of its proper position. 

PARKIA R. Brown 
PARKIA SPECIOSA Hassk. in Flora 25 (1842) Beibl. 55. 
Arbor pete Rumph. Herb. Amb. 3: 51. 

The identification follows Hasskarl, Neue Schliissel (1866) 50, 
which is undoubtedly correct, as proved by the native names 
cited by Rumphius and the indicated uses of the plant. 

CYNOMETRA Linnaeus 
CYNOMETRA CAULI FLORA Linn. Sp. PI. (1753) 382. 
Cynomorium Rumph. Herb. Amb. 1: 163, t. 62. 
Amboina, from cultivated trees in the town of Amboina, Robinson PL 
Rumph. Amb. 530, September 25, 1913, locally known as namu namu. 

* Retained name, Brussels Congress; Gigalobium Boehm. (1760) is older. 

f Contr. U. S. Nat. Herb. 9 (1905) 308. 

$ Philip. Journ. Sci. 5 (1910) Bot. 33; 9 (1914) Bot. 87. 



254 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Cynomorium is one of the few Rumphian species cited by 
Linnaeus in the first edition of the Species Plantarum (1753) 
382, where the reduction to Cynometra cauliflora Linn, was 
made. This is manifestly the correct disposition of it, and 
Linnaeus has been consistently followed by all succeeding 
authors. 

CYNOMETRA RAM I FLORA Linn. Sp. PI. (1753) 382. 

Cynomorium silvestre Rumph. Herb. Amb. 1: 167, t. 63. 
Amboina, Amahoesoe, Robinson PL Rumph. Amb. 531, August 28, 1913, 
on coral limestone cliffs at low altitudes. 

Like Cynomorium, this Rumphian plant was first reduced 
by Linnaeus in the original description of Cynometra ramiflora 
Linn., Sp. PI. (1753) 382. Linnaeus includes in Cynometra 
ramiflora the form with a single pair of leaflets, the Amboina 
plant, and the form with two pairs of leaflets, Iripa of Rheede, 
apparently the form described by Thwaites as Cynometra rami- 
flora Linn. var. heterophylla Thw. The present form, with a 
single pair of leaflets, has very generally been interpreted as 
typical Cynometra ramiflora Linn., and has been indicated by 
Prain as var. genuina, Journ. As. Soc. Beng. 66 2 (1897) 198. 

SIN DORA Miquel 

SINDORA GALEDUPA Prain in Journ. As. Soc. Beng. 66 2 (1897) 483 
(type!). 

Sindora inermis Merr. in Philip. Journ. Sci. 10 (1915) Bot. 314? 
Caju galedupa Rumph. Herb. Amb. 2: 59, t. 13. 

This species is not represented in our Amboina collections. 
The material on which the description was originally based 
was from Celebes and from the small islands of Saleyer and 
Calaua. It grows near the sea and is locally known as caju 
galedupa. Lamarck, Encycl. 2 (1788) 594, cites the Rumphian 
figure and description as a synonym of Galedupa indica, but 
Galedupa indica Lam. as described from Sonnerat's specimens 
is Pongamia glabra Vent.=P. pinnata (Linn.) Merr. Hamilton, 
Mem. Wern. Soc. 6 (1832) 291, thought that it might be a species 
of Copaifera. Wight and Arnott, Prodr. 1 (1834) 262, placed it 
in Pongamia, with doubt. Walpers, Ann. 4 (1857) 581, and 
Miquel, Fl. Ind. Bat. I 1 (1855) 144, erroneously cite it as a 
synonym of Derris forsteniana Blume. Linnaeus, in Stickman 
Herb. Amb. (1754) 9, Amoen. Acad. 4 (1759) 120, placed it 
under Connarus with doubt, but did not refer it to Connarus 
monocarpus Linn, as indicated by Hasskarl, Neue Schlussel 
(1866) 28. Prain has certainly placed it in its correct genus, 



LEGUMINOSAE 



255 



and Sindora galedupa Prain is based wholly on Rumphius. The 
exact status of the species still remains uncertain, but it is 
manifestly very near Sindora coriacea Prain and S. inermis 
Merr. Prain thought that Sindora sumatrana Miq. var. javanica 
Koord. & Valeton might be a synonym of Sindora galedupa. 
I suspect that Sindora inermis Merr. will have to be reduced to 
Sindora galedupa Prain, when botanical material from the Mo- 
luccas is available for comparison. The type of my species was 
from back of the mangrove swamp at Sarangani, southern 
Mindanao, there known as gayugalo, a native name that is 
certainly suggestive of caju galedupa. It differs from the 
Rumphian plant in having six rather than eight leaflets, perhaps 
also in having slightly larger pods, while Rumphius does not 
figure or describe the prominent aciniciform stipules that are 
characteristic of Sindora inermis Merr. 

TAMARINDUS Linnaeus 

TAMARINDUS INDICA Linn. Sp. PL (1753) 34. 
Tamarindus Rumph. Herb. Amb. 2: 90, t. 23. 
Amboina, Robinson PI. Rumph. Amb. 5^9, December 1, 1913, from culti- 
vated trees, locally known as assam. 

This widely distributed and well-known species hardly needs 
discussion. Linnaeus referred the Rumphian figure to his 
species, in Stickman Herb. Amb. (1754) 9, in which he has been 
followed by all succeeding authors. 

I NTS I A Thouars 

INTSIA BIJUGA {Colebr.) O. Kuntze Rev. Gen. PI. 1 (1891) 192. 

Macrolobium bijugum Colebr. in Trans. Linn. Soc. 12 (1819) 359, t. 17. 
Afzelia bijuga A. Gray Bot. Wilkes U. S. Explor. Exped. (1854) 
467, t. 51. 

Outea bijuga DC. Prodr. 2 (1825) 511. 

Intsia amboinensis DC. Prodr. 2 (1825) 509 (type!). 

Macrolobium amboinense Teysm. ex Hassk. in Abh. Naturf. Gesellsch. 

Halle 9 (1866) 189 (type!). 
Metrosideros amboinensis Rumph. Herb. Amb. 3: 21, t. 10. 

This species is not represented in our Amboina collections, 
yet the identity of Metrosideros amboinensis Rumph., at least for 
the most part, with the plant commonly known as Intsia (Afze- 
lia) bijuga 0. Kuntze is certain. Loureiro, Fl. Cochinch. (1790) 
266, cites it with doubt under Baryxylum rufum, but Baryxylum 
rufum Lour, manifestly belongs in the genus Peltophorum; see 
Pierre, Fl. Forest. Cochinch. 4: sub. t. 390. Loureiro's descrip- 
tion of the flowers applies to Peltophorum, of the fruits perhaps 
to Intsia; his specimen in the herbarium of the British Museum 



256 



RUMPHIUS'S HERBARIUM AMBOINENSE 



is a Peltophorum. Possibly more than one species of Intsia are 
included in the description of Metrosideros amboinensis, but 
it is certainly for the most part the common and widely 
distributed Intsia bijuga 0. Kuntze. The figure is poor. The 
description and figure are the whole basis of Intsia amboinensis 
DC. and of Macrolobium amboinense Teysm., the latter not 
appearing in Index Kewensis. 

BAUHINIA Linnaeus 

BAUHINIA LINGUA DC. Prodr. 2 (1825) 516 (type!), excl. syn. Linn. 
Phanera ? lingua Miq. PL Ind. Bat. 1 1 (1855) 67. 
Folium linguae Rumph. Herb. Amb. 5: 1, t. 1. 
Amboina, Soja and Negri lama, Robinson PI. Rumph. Amb. 522, August, 
1913, in fruit, locally known as tabla mulu. 

Folium linguae Rumph. was originally and erroneously re- 
duced by Linnaeus to the Indian Bauhinia scandens Linn., in 
Stickman Herb. Amb. (1754) 18, Amoen. Acad. 4 (1759) 128, 
Syst. ed. 10 (1859) 1015, Sp. PL ed. 2 (1762) 535, in which 
he was followed by numerous authors. Loureiro, however, 
Fl. Cochinch. (1790) 37, cited it under Phanera coccinea 
Lour., the type of which was a Cochin-China plant. Bauhinia 
lingua de Candolle is typified by Folium linguae Rumph., 
the specific name and description being taken wholly from 
Rumphius. Curiously, de Candolle cites as a synonym Bauhinia 
scandens Linn., Sp. PL ed. 1, p. 374, "excl. Rheed, syn./' yet 
Rheede's description and figure are the whole basis of the 
Linnean species, for the first reference given by Linnaeus to Ray 
Suppl. 328 is based wholly on Rheede. 

The Amboina specimen is in fruit, but the species manifestly 
belongs in the section Phanera, as is shown by Rumphius's de- 
scription of the stamens. I have not been able definitely to 
refer the Amboina specimens to any other described species than 
Bauhinia lingua DC, although it seems probable that the same 
form has been described under another specific name or names. 
De Candolle's description calls for a specimen with 3-nerved 
leaf-lobes, as they are thus presented in most of the leaves in 
Rumphius's figure. However, the actual specimens have mostly 
5- or 6-nerved lobes. 

BAUHINIA sp. 

Folium linguae litorea alba Rumph. Herb. Amb. 5: 2. 

This form is not represented in our Amboina collections. 
Hasskarl, Neue Schliissel (1866) 88, suggests that it may be 
Phanera glauca Benth.= Bauhinia glauca Wall., but the cor- 



LEGUMINOSAE 



257 



rectness of this reduction is very improbable. Doubtless a 
future exploration of Amboina will yield material that will 
definitely determine its status. 

D I ALUM Linnaeus 

DIALUM INDUM Linn. Mant. 1 (1767) 24. 

Tamarindus altera Rumph. Herb. Amb. 2: 93. 

From the native names cited by Rumphius, carandje and 
carandjang, and the references cited, this is Dialum indum Linn., 
but the Javan specimens in Rumphius's garden may not have 
been this species. It was from Java, not from Amboina. Ben- 
nett and Miquel both reduce it to Dialum indum Linn. 

CASSIA Linnaeus 

CASSIA MIMOSOIDES Linn. Sp. PI. (1753) 379. 

Amoena moesta Rumph. Herb. Amb. 6: 147, t. 67. f. 1. 
Amboina, Batoe gadjah and Soja road, Robinson PL Rumph. Amb. 535, 
August 5, 1913, on grassy hillsides, altitude 200 to 250 meters. 

Amoena moesta was originally reduced by Linnaeus, in Stick- 
man Herb. Amb. (1754) 28, with doubt, to Cassia procumbens 
Linn., and later, Amoen. Acad. 4 (1759) 135, Syst. ed. 10 
(1759) 1018, to Cassia nictitans Linn., both American species; 
both reductions are erroneous. De Candolle, Prodromus 2 
(1825) 505, reduced it correctly to Cassia angustissima Lam., 
but Lamarck's species is a synonym of Cassia mimosoides Linn, 
or at most represents merely a variety of it, C. mimosoides 
Linn, var angustissima (Lam.) Walp. 

CASSIA ALATA Linn. Sp. PL (1753) 378. 

Herpetica alata Raf. Sylv. Tellur. (1838) 123. 

Cassia alata Linn. var. rumphiana DC. Prodr. 2 (1825) 492 (type!). 
Herpetica Rumph. Herb. Amb. 7: 35, t. 18. 

Amboina, in a sago swamp near the town of Amboina, Robinson PL 
Rumph. Amb. 546, July 25, 1913. 

The reduction of Herpetica to Cassia alata was originally made 
by Linnaeus, Amoen. Acad. 4 (1759) 136, is the correct disposi- 
tion of it, and has been accepted by all authors. 

CASSIA TORA Linn. Sp. PL (1753) 376. 

Cassia obtusifolia Linn. Sp. PL (1753) 377. 

Gallinaria rotundifoiia Rumph. Herb. Amb. 5: 283, t. 97, f. 2. 
Amboina, Batoe merah, Robinson PL Rumph. Aw6. 525, August 15, 1913, 
at low altitudes near the seashore ; also Rel. Robins. 2480 from Boeton Island, 
July 13, 1913. 

A weed of pantropic distribution, originating in tropical 

144971 17 



258 



RUMPHIUS'S HERBARIUM AMBOIfcENSE 



America. Linnaeus reduced the Rumphian figure to both his 
Cassia tora and Cassia obtusifolia, the latter being a synonym of 
the former. It is considered as Cassia tora Linn., in Stickman 
Herb. Amb. (1754) 22, Amoen. Acad. 4 (1759) 131, Syst. ed. 
10 (1759) 1016, and as Cassia obtusifolia Linn, in his Species 
Plantarum, ed. 2 (1762) 539. The early authors, following 
Linnaeus for the most part, considered it as Cassia obtusifolia 
Linn., but practically all recent authors have properly placed 
it under Cassia tora Linn. 

CASSIA OCCIDENTALIS Linn. Sp. PI. (1753) 377. 

Gallinaria acutifolia Rumph. Herb. Amb. 5: 283, t. 97, f. 1. 
Amboina, along the beach near the town of Amboina, Robinson PI. 
Rumph. Amb. 53U, August 8, 1913; also represented by Rel. Robins. 2519 
from Bali Island, July 7, 1913. 

Linnaeus originally reduced Gallinaria acutifolia to Cassia 
sophera Linn., in Stickman Herb. Amb. (1754) 22, Amoen. 
Acad. 4 (1759) 131, Syst. ed. 10 (1759) 1017, Sp. PL ed. 2 
(1762) 542, in which he was consistently followed by all early 
authors. Vogel, Hasskarl, and Miquel, however, have correctly 
placed it as Cassia occidentalis Linn., which, like the preceding 
species, is a pantropic weed of American origin. 

CASSIA GLAUCA Lam. Encycl. 1 (1785) 647. 
Fios flavus Rumph. Herb. Amb. 4: 63, t. 23. 

This species is not represented in our Amboina collections. 
The figure, however, is unmistakably that of Cassia glauca Lam., 
where it was definitely placed by Miquel and by Hasskarl. Bur- 
man f., Fl. Ind. (1768) 96, erroneously reduced it to Cassia plani- 
siliqua, or at least Cassia planisiliqua Burm. f. is entirely dif- 
ferent from Cassia planisiliqua Linn. Lamarck, Encycl. 1 
(1785) 644, suggests its comparison with Cassia chinensis Lam., 
but it certainly is not this, although the exact status of Cassia 
chinensis Lam. is doubtful; from the description it may be 
Cassia occidentalis Linn., although Bentham thought it might 
be Cassia sophera Linn. 

CASSIA SOPHERA Linn. Sp. PI. (1753) 379. 
Soffera Rumph. Herb. Amb. 4: 55. 

This species is not represented in our Amboina collections. 
The plant described by Rumphius is probably the common and 
widely distributed Cassia sophera Linn., as indicated by Hass- 
karl, Neue Schliissel (1866) 75. 



LEGUMINOSAE 



259 



CASSIA FISTULA Linn. Sp. PL (1753) 377. 

Cathartocarpus fistula Pers. Syn. 1 (1805) 459. 
Cassia fistula Rumph. Herb. Amb. 2: 83, t. 21. 

This common species is not represented in our Amboina col- 
lections, but it is so well known that it hardly needs discussion. 
Rumphius's plate is good. It was first reduced by Linnaeus, 
in Stickman Herb. Amb. (1754) 9, and has been consistently 
cited under Cassia fistula Linn, by all subsequent authors. 

CASSIA JAVANICA Linn. Sp. PI. (1753) 379. 

Carina fistula javanica Rumph. Herb. Amb. 2: 86. 
Cassia fistula silvestris Rumph. Herb. Amb. 2: 88, t. 22. 

This species is not represented in our Amboina collections. 
There is very little doubt, however, that the first of the Rumphian 
names cited above is referable to Cassia javanica, although 
Cassia fistula silvestris may include two different species. The 
reduction of t. 22 to Cassia javanica Linn, was first made by 
Lamarck, Encycl. 1 (1785) 649, and has been very generally 
accepted. Roxburgh, Fl. Ind. ed. 2, 2 (1832) 337, referred it 
to Cassia bacillus Gaertn., which, however, is generally con- 
sidered to be a synonym of Cassia javanica Linn. Other names 
involved in the reduction are Cassia marginata Roxb. and C. 
nodosa Ham. 

The plants briefly discussed by Rumphius, op. cit. 89, under the 
names bilalangh, cajudju, and ke ule are indeterminable from 
data at present available. Hasskarl, Neue Schlussel (1866) 30, 
has suggested that the second one may be a Pterocarpus and 
that the other two may be referable to Cassia. The only logical 
method of determining these forms is to carry on field work 
with special reference to the native names. 

CAESALPI N I A Linnaeus 

CAESALPI N I A SAPPAN Linn. Sp. PI. (1753) 381. 

Lignum sappan Rumph. Herb. Amb. 4: 56, t. 21. 

Amboina, Wae, Robinson PI. Rumph. Amb. 568, November 26, 1913, 
cultivated, locally known as lolan tuni. 

This well-known species hardly needs discussion. Lignum 
sappan was originally reduced to Caesalpinia sappan Linn, by 
Linnaeus, in Stickman Herb. Amb. (1754) 15, and has been 
consistently so cited by all other authors who have had occasion 
to quote Rumphius. 



260 



RUMPHIUS'S HERBARIUM AMBOINENSE 



CAESALPINIA PU LCH ERRI M A (Linn.) Sw. Obs. (1791) 166. 
Poinciana pulcherrima Linn. Sp. PL (1753) 380. 
Crista pavonis Rumph. Herb. Amb. 4: 53, t. 20. 

Amboina, Robinson PL Rumph. Amb. 51>2, September, 1913, from culti- 
vated plants in the town of Amboina, including both the red and yellow 
and the yellow-flowered forms. 

This commonly cultivated plant, of tropical American origin, 
is well figured by Rumphius. Crista pavonis was first reduced 
by Linnaeus, to Poinciana pulcherrima, in Stickman Herb. Amb. 
(1754) 15, Amoen. Acad. 4 (1759) 126, Syst. ed. 10 (1759) 
1018; but in the second edition of his Species Plantarum (1762) 
544, he erroneously referred it to Poinciana bijuga Linn. 

CAESALPINIA CRISTA Linn. Sp. PI. (1753) 380, pro majore parte, excl. 
Fl. Zeyl. 157 quoad Herm. zeyl. 12. 

Guilandina bonducella Linn. Sp. PI. ed. 2 (1762) 545. 
Guilandina bonduc Linn. Sp. PL (1753) 381, p. p., excl. Fl. Zeyl. 156. 
Caesalpinia bonducella Flem. Asiat. Res. 11 (1810) 159. 
Guilandina crista Small Fl. Southeast. U. S. (1905) 591. 
Guilandina bonduc Linn. var. minus DC. Prodr. 2 (1825) 480. 
Caesalpinia jayabo var cyanosperma Maza in Anal. Soc. Esp. Hist, 
j Nat. 19 (1890) 234. 

Frutex globulorum majorum Rumph. Herb. Amb. 5: 92, t. 49, /. 1. 

The widely distributed plant commonly known as Caesalpinia 
bonducella Flem. is not represented in our Amboina collections. 
There is no question whatever as to the identity of the plant 
Rumphius figures, as his illustration is an excellent one. The 
synonymy, like that of Caesalpinia jayabo Maza (C. bonduc 
auctt.), is exceedingly complicated and was first adjusted by 
Urban, Symb. Antil. 2 (1900) 269. I agree with Doctor Urban 
in the application of the Linnean name in spite of the fact 
that the first reference given by Linnaeus in the original de- 
scription of the species applies to the plant commonly known as 
Caesalpinia nuga (Linn.) Ait., as originally pointed out by 
Trimen, Fl. Ceyl. 2 (1894) 99; see under Caesalpinia nuga Ait., 
infra, page 261. Skeels, Science N. S. 37 (1913) 921, would 
interpret Fl. Zeyl. 157 strictly on the basis of Hermann zeyl. 
12, as the type of Caesalpinia crista Linn., thus making the 
species exactly the same as Caesalpinia nuga (Linn.) Ait., re- 
ducing the latter as a synonym. At the same time he would 
interpret Fl. Zeyl. 156 as the type of Guilandina bonduc Linn., 
thus making Caesalpinia bonduc the proper name for the plant 
described in most botanical works as Caesalpinia bonducella 
Flem. 



LEGUMINOSAE 



261 



CA ESALPI N I A JAYABO Maza in Anal. Soc. Esp. Hist. Nat. 19 (1890) 234. 
Guilandina bonduc Linn. Sp. PI. (1753) 381, p. p., quoad Fl. Zeyl. 156. 
Caesalpinia bonduc auctt. 

Guilandina bonduc Linn. var. majus DC. Prodr. 2 (1825) 480. 
Guilandina major Small Fl. Southeast. U. S. (1903) 591. 
Caesalpinia glabra Merr. in Philip. Journ. Sci. 5 (1910) Bot. 54, non 

Guilandina glabra Mill. 
Frutex globulorum femina Rumph. Herb. Amb. 5: 89, t. 48. 

Amboina, Wae, Robinson PL Rumph. Amb. 564, November 25, 1913, climb- 
ing over trees at low altitudes. 

The reduction of Frutex globulorum femina was first made by 
Linnaeus, who placed it as a synonym of Guilandina bonduc 
Linn, in his Species Plantarum ed. 2 (1762) 545, and most 
authors have been content to accept this name as the proper 
one for the species. The synonymy is very complicated, but 
Guilandina bonduc Linn, as originally published by Linnaeus in 
1753 is for the most part identical with G. bonducella Linn, 
as published by Linnaeus in the year 1762, and both are, for 
the most part, the same as Caesalpinia crista Linn. 1753. The 
synonymy has been very fully discussed by Urban, Symb. Antil. 
2 (1900) 269, 272. Guilandina bonduc Linn. var. majus DC. 
was based on the Linnean description, as given in the second 
edition of the Species Plantarum, and the figure of Rumphius 
cited above and in turn is the basis of Guilandina major Small. 
Guilandina glabra Mill, is an entirely different species, if Miller's 
short description be correct, and I erred in 1910 in proposing to 
take up this specific name. It is apparent that Maza's specific 
name Caesalpinia jayabo is the earliest valid one for this pan- 
tropic plant. Caesalpinia bonduc Roxb., based on Guilandina 
bonduc Linn., is certainly untenable and should be abandoned. 

CAESALPINIA NUGA (Linn.) Ait. Hort. Kew. ed. 2, 3 (1811) 32. 
Guilandina nuga Linn. Sp. PI. ed. 2 (1762) 546 (type!). 
Caesalpinia crista Linn. Sp. PI. (1753) 380, pro minore parte, FL Zeyl. 

157 quoad Herm. zeyl. 12. 
Nugae silvarum litoreae et terrestres Rumph. Herb. Amb. 5: 94, t. 50. 
Amboina, Eri, Robinson PL Rumph. Amb. 539, September 22, 1913, along 
the seashore; Amahoesoe, Robinson PL Rumph. Amb. 538, August 13, 1913, 
in thickets at an altitude of about 30 meters, locally known as galachi and 
pohon baduri. ,. 

The Rumphian reference is the whole basis of Guilandina 
nuga Lmn.=Caesalpinia nuga (Linn.) Ait. and has been inter- 
preted as such by all authors. In this connection it is to be 
noted that Caesalpinia crista Linn., Sp. PI. (1753) 380, is, in 
part, the same as Caesalpinia nuga Ait. The first reference 
in the original description of the species is to Fl. Zeyl. No. 157, 



262 RUMPHIUS'S HERBARIUM AMBOINENSE 

and Hermann's specimen is not Caesalpinia crista Linn, as 
currently interpreted, but is C. nuga Ait. ; see Trimen, Fl. Ceyl. 
2 (1894) 99. However, the other references in the Species 
Plantarum and the first reference in the Flora Zeylanica, Pluk. 
Aim. U. t. 2 f. 2, are apparently Caesalpinia crista Linn, as 
generally understood, so that Caesalpinia crista Linn, is here 
maintained for the plant so described by all authors; see 
page 260. 

ORMOSIA * Jackson 

ORMOSIA CALAVENSIS Azaola ex Blanco Fl. Filip. ed. 2 (1845) 230. 
Pongamia (?) corallaria Miq. Fl. Ind. Bat. 1 1 (1855) 149 (type!). 
Corallaria latifolia Rumph. Herb. Amb. 3: 175, t. 110. 

This characteristic species is not represented in our Amboina 
collections. The reduction to Ormosia calavensis Azaola is made 
after a careful study of Rumphius's figure and description and 
a comparison with a very full series of specimens from northern 
Luzon to southern Mindanao and from the Palau Islands. 
Corallaria latifolia is the whole basis of Pongamia corallaria 
Miq., so this specific name is available should future botanical 
exploration of the Moluccas yield material that shows the plant 
Rumphius described to be specifically distinct from that of the 
Philippine and the Palau Islands. Teysmann, quoted by 
Hasskarl, Neue Schlussel (1866) 63, referred it to the genus 
Macropteris, this being a synonym of Ormosia. 

SOPH OR A Linnaeus 

SOPHORA TOM ENTOSA Linn. Sp. PL (1753) 373. 
Anticholerica Rumph. Herb. Amb. 4: 60. t. 22. 
Amboina, Paso, Robinson PI. Rumph. Amb. 5J>5, October 31, 1913, along 
the seashore. 

Anticholerica of Rumphius has been confused by most authors, 
following Linnaeus, with Sophora heptaphylla Linn. It is 
typical Sophora tomentosa Linn. It was originally reduced 
by Linnaeus to Sophora heptaphylla, in Stickman Herb. Amb. 
(1754) 16, Amoen. Acad. 4 (1759) 126, Syst. ed. 10 (1759) 
1015, Sp. PI. ed. 2 (1762) 533, in which he was followed by all 
authors up to 1825, when de Candolle, Prodromus 2 : 96, suggested 
that" the Rumphian figure referred to Sophora tomentosa Linn. 
Sophora heptaphylla Linn, is in itself a mixture. The type is 
FL Zeyl. 10U, and Hermann's specimen on which it is based is 
Derris sinuata Benth.-^Derris heptaphylla (Linn.) Merr., 
page 273. 



* Retained name, Vienna Code; Toulichiba Adans. (1763) is older. 



LEGUMINOSAE 



263 



CROTALARIA Linnaeus 

CROTALARIA RETUSA Linn. Sp. PL (1753) 715. 

Crotalaria I major Rumph. Herb. Amb. 5: 278, t. 96, f. I. 
Amboina, Hatiwe, Robinson PI. Rumph. Amb. 5^7, September 4, 1913, 
near the seashore; also represented by Rel. Robins. 2^6 A from Macassar, 
Celebes. 

The reduction of the form of Crotalaria figured by Rumphius 
to Crotalaria retusa Linn, was first made by Linnaeus, in Stick- 
man Herb. Amb. (1754) 21, Amoen. Acad. 4 (1759) 130, Syst. 
ed. 10 (1759) 1159, Sp. PL ed. 2 (1763) 1004, and is certainly 
the correct disposition of it. 

CROTALARIA QU I NQU EFOLI A Linn. Sp. PI. (1753) 716. 
Crotalaria II minor Rumph. Herb. Amb. 6: 278. 

This species is not represented in our Amboina collections. 
The reduction of the Rumphian description follows Hasskarl, 
Neue Schlussel (1866) 121, which is manifestly the correct 
disposition of it. 

CROTALARIA CHINENSIS Linn. Syst. ed. 10 (1759) 1158. 
Crotalaria III agrestis Rumph. Herb. Amb. 5: 297. 
Amboina, Soja road, Robinson PI. Rumph Amb. J+68, August 1, 1913, on 
grassy hillsides, altitude about 200 meters. 

The identity of the recently collected material with Crotalaria 
III agrestis is somewhat doubtful, as Rumphius's description is 
short and imperfect. It cannot possibly be Crotalaria verrucosa 
Linn., as placed by Hasskarl, Neue Schlussel (1866) 121. 

CROTALARIA LI N I FOLIA Linn. f. Suppl. (1781) 322. 

Phaseolus montanus III Rumph. Herb. Amb. 6: 146. 

This suggested identification of Phaseolus montanus III is 
scarcely more than a guess, but the description, with a slight 
modification of the Latin translation, seems to apply very closely 
to the form of Crotalaria linifolia Linn. f. described by Vogel 
as C. stenophylla and by Matsumura as C. formosana. In the 
Dutch description the fruits are described as "met twee rug- 
gens," which I translate "with two ridges." The Latin descrip- 
tion reads "cum binis alis," translated by Hasskarl as "bialatis." 
With the modification of the description suggested by translating 
"ruggens" as "ridges" instead of as "wings," there is nothing 
in the entire description that does not apply to Crotalaria steno- 
phylla Vogel. 

Phaseolus montanus IV Rumph. 1. c. is apparently but a dwarfed 
form of Crotalaria linifolia Linn. f. 



264 



RUMPHIUS'S HERBARIUM AMBOINENSE 



I N DIGOFERA Linnaeus 

I N DIGOFERA T! NCTORI A Linn. Sp. PI. (1753) 751. 

Indicum Rumph. Herb. Amb. 5: 220, quoad descr. (excl. t. 80—1. 

suffruticosa Mill.). 

Probably both Indigo f era tinctoria Linn, and L suffruticosa 
Mill. (I. anil Linn.) are included in the Rump hi an discussion of 
Indicum, but the description of the pods as "digiti articulum 
longae" applies unmistakably to Indigofera tinctoria Linn., not 
to /. suffruticosa Mill. The figure, however, unmistakably repre- 
sents Indigofera suffruticosa Mill., as shown by the relatively 
short, strongly curved pods. Indicum was reduced by Linnaeus 
to Indigofera tinctoria Linn., in Stickman Herb. Amb. (1754) 
21, Amoen. Acad. 4 (1759) 130, Syst. ed. 10 (1759) 1171, Sp. 
PL ed. 2 (1763) 1061, in which he has been followed by most 
authors. Lamarck, Encycl. 3 (1789) 244, referred it to Indigo- 
fera anil Linn.=7. suffruticosa Mill., which is the correct disposi- 
tion of the figure. Other authors have referred it to Indigofera 
anil Linn. var. orthocarpa DC, /. tinctoria Linn. var. macrocarpa 
DC, and /. tinctoria Linn. var. brachycarpa DC 

INDIGOFERA SUFFRUTICOSA Mill. Gard. Diet. ed. 8 (1768) no. 2. 
Indigofera anil Linn. Mant. 2 (1771) 272. 
indicum Rumph. Herb. Amb. 5: quoad t, 80. 
Amboina, near the town of Amboina, Robinson PI. Rumph. Amb. 195, 
August 8, 1913, near the seashore. 

This is the form figured by Rumphius, but so far as the 
description, goes the essential characters by which Indigofera 
tinctoria Linn, and /. suffruticosa Mill, are distinguished are 
those of the former species. As indicated above probably both 
are included in the general discussion of Indicum. 

The several forms briefly described under Indicum I consider 
to be indeterminable, although Hasskarl, Neue Schliissel (1866) 
113, indicates their possible positions as follows : Indicum silvestre 
Rumph. Herb. Amb. 5: 222=Indigofera sp.?; Indicum silvestre e 
Madagascar Rumph. 1. c. 22S=Indigofera linifolia Retz. var. an- 
gustissima Miq.?; Indicum brasilianum Rumph. 1. c. 22A=Indigo- 
fera sp.?; Indicum spurium Rumph. 1. c. 224= Indigofera celebica 
Miq.?. 

TEPHROSIA* Persoon 
TEPHROSIA sp. aff. T. purpurea Pers. 

Phaseolus montanus I Rumph. Herb. Amb. 6: 146. 

No representative of the genus Tephrosia occurs in our Am- 
boina collections. The plant described by Rumphius, however, 

* Retained name, Vienna Code; Cracca Linn. (1763) is older. 



LEGUMINOSAE 



265 



is manifestly a Tephrosia, and it is probably a form of the 
widely distributed T. purpurea Pers. Hasskarl, Neue Schlussel 
(1866) 176, considers that it represents a species near Tephrosia 
timoriensis DC, but de Candolle's species is generally considered 
to be a synonym of T. purpurea Pers. 

Phaseolus moxitanus alter Rumph., 1. c. 146, may also represent 
a species of Tephrosia. The description, however, is too short 
and imperfect to warrant an identification of it with the data 
now available. 

SESBAN I A * Persoon 

SESBANIA SESBAN (Linn.) Merr. in Philip. Journ. Sci. 7 (1912) Bot. 
235. 

Aeschynomene sesban Linn. Sp. PL (1753) 714. 
Coronilla sesban Willd. Sp. PI. 3 (1806) 1147. 
Sesban aegyptiacus Poir. in Lam. Encycl. 7 (1806) 128 (type!). 
Aeschynomene moluccana Kostel. Allg. Med.-Pharm. Flora 4 (1835) 
1285. 

Emerus sesban O. Kuntze Rev. Gen. PI. 1 (1891) 180. 
Gajatus niger Rumph. Herb. Amb. 4: 64, t. 2U- 

This species is not represented in our Amboina collections. 
The Rumphian figure was erroneously reduced by Linneaus to 
Aeschynomene indica Linn., in Stickman Herb. Amb. (1754) 
16, Amoen. Acad. 4 (1759) 126, Syst. ed. 10 (1759) 1158, Sp. 
PI. ed. 2 (1763) 1061, in which he was generally followed by 
the older authors. Wight and Arnott, Prodr. (1834) 219, placed 
it under Aeschynomene roxburghii Spreng., a synonym of A. 
indica Linn. It is apparently the type of Aeschynomene moluc- 
cana Kostel. (publication not seen by me). Miquel, Fl. Ind. 
Bat. I 1 (1855) 287, thought it might be referable to Sesbania 
cochinchinensis (Lour.) DC, which is probably a synonym of 
Sesbania sesban (Linn.) Merr. Teysmann correctly placed it 
as a synonym of Sesbania aegyptiaca Pers., which manifestly is 
a synonym of Sesbania sesban (Linn.) Merr. From Rumphius's 
description it is exactly the form described by Wight and Arnott, 
Prodr. (1834) 214, as Sesbania aegyptiaca Pers. var. bicolor 
W. & A.f 

SESBANIA CAN N ABI N A (Retz.) Pers. Syn. 2 (1807) 316. 
Aeschynomene cannabina Retz. Obs. 5 (1789) 26. 
Agati cannabina Desv. Journ. Bot. 1 (1813) 120. 
Gajatus luteus Rumph. Herb. Amb. 4: 64. 

This species is not represented in our Amboina collections. 

* Retained name, Brussels Congress; Sesban Adans. and Agati Adans. 
(1763) are older. 

f See Prain in Journ. As. Soc. Beng. 66 2 (1897) 367. 



266 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Hasskarl, Neue Schlussel (1866) 76, discusses it under Aeschy- 
nomene indica Linn. var. aspera Hassk. and under Sesbania 
polyphylla Miq. The Rumphian description calls for a plant 
with larger leaves and longer pods than Gajatus niger (Sesbania 
sesban Merr.) and with yellow flowers, and I believe that the 
plant described is unmistakably Sesbania cannabina (Retz.) 
Pers. It was introduced into Amboina in Rumphius's time. 
The form from Bali with white flowers, merely mentioned by 
Rumphius, is indeterminable, the only character given being 
that the flowers are white. 

SESBANIA GRANDI FLORA (Linn.) Pers. Syn. 2 (1807) 316. 

Robinia grandifiora Linn. Sp. PI. (1753) 722. 

Aeschynomene grandifiora Linn. Sp. PI. ed. 2 (1763) 1060. 

Coronilla grandifiora Willd. Sp. PI. 3 (1800) 1145. 

Agati grandifiora Desv. in Journ. Bot. 1 (1813) 120, t. h, /• 6. 

Turia Rumph. Herb. Amb. 1 : 188, t. 76. 

Turia minor Rumph. Herb. Amb. 1: 190, t. 77. 

Amboina, Binting, Robinson PL Rumph. Amb. 532, September 27, 1913, 
at low altitudes, locally known as turi. 

Turia was originally reduced by Linnaeus to Aeschynomene 
grandifiora Linn, in his Species Plantarum, ed. 2 (1763) 1060, 
and the figure has very generally been cited by various authors 
under one or the other of the synonyms cited above. Turia 
minor Rumph. is the form with reddish or purplish flowers, 
considered by early authors to represent a distinct species, 
Sesbania coccinea (Linn, f.) Pers. (Aeschynomene coccinea Linn, 
f., Coronilla coccinea Willd., Sesban coccinea Poir., Agati cocci- 
nea DC), and by others considered as merely a variety of 
Sesbania grandifiora Pers. It is manifestly but a color form of 
the common and widely distributed Sesbania grandifiora (Linn.) 
Pers. 

ORMOCARPUM * de Candolle 

ORMOCARPUM ORIENTALE (Spreng.) comb. nov. 

Parkinsonia orientalis Spreng. Syst. 4 (1827) Cur. Post. 170 (type!). 
Ormocarpum glabrum Teysm. & Binn. in Tijschr. Ned. Ind. 27 (1854) 
56. 

Solulus arbor Rumph. Herb. Amb. 3: 200, t. 128. 

This species is not represented in our Amboina collections, but 
according to Rumphius the plant is not a native of Amboina, 
occurring there as an introduced and planted one. Loureiro, Fl. 
Cochinch. (1790) 454, cites Solulus arbor as representing his 

Diphaca cochinchinensis=Ormocarpum cochinchinense (Lour.) 



* Retained name, Vienna Code; Diphaca Lour. (1790) is older. 



LEGUMINOSAE 



267 



Merr. (0. sennoides DC.), but the Rumphian figure and descrip- 
tion manifestly apply to Ormocarpum glabrum T. & B. The 
oldest specific name, however, if this form be maintained as 
distinct from Ormocarpum cochinchinense, is that supplied by 
Parkinsonia orientate Spreng., which was based wholly on the 
Rumphian figure and description. 

ARACHIS Linnaeus 

ARACHIS HYPOGAEA Linn. Sp. PI. (1753) 741. 

Chamaebalanus japonica Rumph. Herb. Amb. 5: 426, t. 156, f. 2. 

The common peanut is not represented in our Amboina col- 
lections, although doubtless it is still cultivated there as it is 
in most warm countries. The reduction of the Rumphian figure 
was first made by Linnaeus, in Stickman Herb. Amb. (1754) 24, 
Amoen. Acad. 4 (1759) 132, Syst. ed. 10 (1759) 1167, which has 
been followed by all authors except Loureiro, who proposed to 
call it Arachis asiatica Lour., Fl. Cochinch. (1790) 430. Lou- 
reiro's species is manifestly a synonym of Arachis hypogaea 
Linn. 

DESMODIUM * Desvaux 

DESMODIUM ORMOCARPOIDES DC. Prodr. 2 (1825) 327, non auct. plur. 
Desmodium ormocarpoides Desv. in Mem. Soc. Linn. Paris 1825 (1826) 

307, non auct. plur. 
Hedysarum adhaerens Poir. in Lam. Encycl. Suppl. 5 (1817) 15, non 
Vahl. 

Desmodium dependens Blume ex Miq. Fl. Ind. Bat. 1 1 (1855) 248. 
Phaseolus montanus Rumph. Herb. Amb. 6: quoad t. 66 p. p. (excl. 
descr. ! ) . 

Amboina, Way tommo, Robinson PL Rumph. Amb. 555, August 16, 1913. 
in light woods at low altitudes; Amahoesoe, PI. Rumph. Amb. 556, August 
30, 1913, at low altitudes, locally known as rumpit makal. 

The plant figured by Rumphius does not agree with any of 
the eight forms described under the name Phaseolus montanus. 
Burman f., Fl. Ind. (1768) 164, referred the figure to Hedysarum 
gangeticum Linn. =Desm,odium gangeticum (Linn.) DC, in 
which he was followed by numerous authors; this reduction, 
however, is entirely wrong. The figure is, for the most part, 
an excellent representation of Desmodium dependens Blume, 
which was originally described from specimens originating in 
the Moluccas and in New Guinea, which Gagnepain, Not. Syst. 
3 (1916) 256, has recently shown to be exactly the form described 
by de Candolle and by Desvaux as Desmodium ormocarpoides, 

* Retained name, Vienna Code; Meibomia Adans. (1763) is older. 



268 



RUMPHIUS'S HERBARIUM AMBOINENSE 



a species that has consistently been misinterpreted by modern 
authors on account of de Candolle's original insufficient descrip- 
tion. The drawing apparently represents two species. The 
infructescence and fruits shown on the right-hand branch are 
distinctly different from those shown on the left-hand branch; 
the former is apparently Desmodium ormocarpoides auct., non 
BC.=Desmodium zonatum Miq.,* and the latter is typical D. 
ormocarpoides DC. (D. dependens Blume) . 

DESMODIUM UMBELLATUM (Linn.) DC. Prodr. 2 (1825) 325. 
Hedysarum umbellatum Linn. Sp. PL (1753) 747. 
Aeschynomene arborea Linn. Sp. PI. (1753) 713, non Desmodium 
arboreum Sweet. 

Dendrolobium umbellatum W. & A. ex Benth. Pi. Jungh. (1852) 216. 
Meibomia umbellata 0. Kuntze Rev. Gen. PI. 1 (1891) 197. 
Folium crocodili latifoiium Rumph. Herb. Amb. 4: 112, t. 52. 
Folium crocodili parvifollum Rumph. 1. c. 113. 

Amboina, Paso, Robinson PL Rumph. Amb. 560, October 21, 1913, along 
the seashore; Ayer putri, Robinson PL Rumph. Amb. 559, July 28, 1913. 

Folium crocodili Rumph. was originally reduced by Linnaeus 
to Hedysarum umbellatum Linn., in Stickman Herb. Amb. 
(1754) 17, Amoen. Acad. 4 (1759) 127, Syst. ed. 10 (1759) 
1170, Sp. PL ed. 2 (1763) 1053, which, as Desmodium umbella- 
tum (Linn.) DC, is certainly the correct dispositon of it. 
Miquel, FL Ind. Bat. I 1 (1855) 263, erroneously placed it under 
Dendrolobium cephalotes Benth. The two forms described by 
Rumphius as latifoiium and parvifolium are manifestly referable 
to the same species. 

DESMODIUM TRIQUETRUM (Linn.) DC. Prodr. 2 (1825) 326. 

Hedysarum triquetrum Linn. Sp. PI. (1753) 746. 
Pteroloma triquetrum Benth. in Miq. PI. Jungh. (1852) 220. 
Meibomia triquetra O. Kuntze Rev. Gen. PI. 1 (1891) 197. 
Phaseolus montanus VI, VII Rumph. Herb. Amb. 6: 146. 

Amboina, Soja road, Robinson PL Rumph. Amb. 557, August 1, 1913, 
on hillsides, altitude 200 to 300 meters. 

The reduction of Phaseolus montanus VI to Hedysarum tri- 
quetrum Linn, was first made by Loureiro, FL Cochinch. (1790) 
448, which, as Desmodium triquetrum (Linn.) DC, is certainly 
the correct disposition of it. Phaseolus montanus VII was first 

* At the time of publication of this the continuation of Gagnepain's paper 
in Not. Syst. 3, no. 9, has not reached me, so I am not certain that Des- 
modium zonatum Miq. is the name selected by him for Desmodium ormo- 
carpoides auct., non DC. 



LEGUMINOSAE 



269 



reduced here by Linnaeus, Sp. PL ed. 2 (1763) 1052, which is 
probably the correct dispositon of it, although not certain. 

DESMODIUM GANGETICUM (Linn.) DC. Prodr. 2 (1825) 327. 
Hedysarum gangeticum Linn. Sp. PL (1753) 746. 
Meibomia gangetica O. Kuntze Rev. Gen. PI. 1 (1891) 196. 
Crotalaria montana V Rumph. Herb. Amb. 6: 146? (haud t. 66 quae 
est Desmodium ormocarpoides DC. et D. zonatum Miq.). 
Amboina, Batoe gadjah, Robinson PL Rumph. Amb. 558, August 5, 1913, 
on grassy hillsides, altitude about 150 meters. Also represented by Rel. 
Robins. 2537 from Bali Island, July 7, 1913. ' 

The reduction of Crotalaria montana V to Desmodium gan- 
geticum (Linn.) DC. is rather unsatisfactory, but follows Bur- 
man f., Loureiro, Poiret, de Candolle, and other authors. All, 
however, apparently based their conception of the Rumphian 
plant chiefly if not wholly upon the figure indicated by Rumphius 
as Phaseolus montanus. None of the eight forms actually de- 
scribed by Rumphius under the heading Phaseolus montanus 
agrees with the figure, which unmistakably is Desmodium or- 
mocarpoides DC. ; see page 267. The Rumphian description 
of Crotalaria montana V applies to Desmodium gangeticum DC. 
sufficiently closely except in the description of the leaves as "semi 
digitum longa, ac pennam lata." 

Phaseolus montanus VIII Rumph., Herb. Amb. 6: 146, is inde- 
terminable from any data at present available. It was from 
Macassar, Celebes, where it was known as tsjeme tsjeme. Hass- 
karl, Neue Schlussel (1866) 176, suggests that it may be a 
species of Sophora near S. glabra Hassk. 

PSEUDARTHRIA Wight and Arnott 

PSEUDARTHRIA VISCIDA (Linn.) W. & A. Prodr. (1834) 209. 

Hedysarum viscidum Linn. Sp. PI. (1753) 747. 

Desmodium viscidum DC. Prodr. 2 (1825) 336. 

Desmodium timoriense DC. 1. c. 327. 

Phaseolus adhaerens Rumph. Herb. Amb. 6: 150. 
Amboina, Amahoesoe, Robinson PL Rumph. Amb. 55J+, August 30, 1913, 
at low altitudes, locally known as bunga pasang pasang. Also represented 
by Rel. Robins. 24-99 from Boeton, and Rel. Robins. 2536 from Bali, July, 
1913. 

The identification of Phaseolus adhaerens Rumph. with PseUd- 
arthria viscida W. & A. is not certain. The Rumphian plant 
may be a Desmodium rather than a Pseudarthria. Hasskarl, 
Neue Schlussel (1866) 177, considers that it is Desmodium stipu- 
laceum DC. var. aparine (Hassk.) Miq. Fl. Ind. Bat. I 1 (1855) 
252. 



270 RUMPHIUS'S HERBARIUM AMBOINENSE 

f 

DALBERGIA* Linnaeus f. 

DALBERGIA PARVI FLORA Roxb. Hort. Beng. (1814) 98, nomen nudum, 
Fl. Ind. ed. 2, 3 (1832) 225. 
Dalbergia Zollinger 'iana Miq. Fl. Ind. Bat. 1 1 (1855) 130. 
Lacca lignum Rumph. Herb. Amb. 5 : 17, t. 13. 

This species is not represented in our Amboina collections. 
The reduction of Lacca lignum to Dalbergia parviflora Roxb. is 
certainly correct and was first suggested, as D. Zollinger 'iana 
Miq., by Teysmann in a letter to Hasskarl, cited by Hasskarl, 
Neue Schlussel (1866) 90.f 

This species of Dalbergia extends from Indo-China and the 
Malay Peninsula to Sumatra, Borneo, Java, Celebes, Halmaheira, 
and Amboina. 

The forms briefly described by Rumphius as Lacca lignum 
ruffum, Herb. Amb. 5 : 18, L. lignum femina, 1. c. 20, and L. lignum 
e Java, 1. c. 20, are indeterminable from data now available 
and may be referable to Dalbergia or to other genera of the 
Leguminosae, or they may even belong in other families. 

PTEROCARPUS Linnaeus 

PTEROCARPUS INDICUS Willd. Sp. PI. 3 s (1800) 904 (type!). 
Lingoum rubrum Rumph. Herb. Amb. 2: 205, t. 70. 
Amboina, Hoenoet, Robinson PI. Rumph. Amb. 550, October 18, 1913, 
in remnants of forests, altitude 200 meters, locally known as kayu lingua. 

This was originally reduced by Linnaeus merely to the genus 
Pterocarpus, in Stickman Herb. Amb. (1754) 10, but in his 
Species Plantarum, ed. 2 (1763) 1662, he erroneously placed it 
as a synonym of Pterocarpus draco Linn. It is the entire basis 
of Pterocarpus indicus Willd., which species must be interpreted 
wholly from the Rumphian figure and description. The species 
is widely distributed in the Malayan region and very generally 
has been correctly interpreted by modern botanists, as Rumph- 
ius's figure of the plant is an excellent one. 

Probably referable here are the forms described by Rumphius, 
1. c. 206, 209, 210, as Lingoum II album and III rubrum. 

PTEROCARPUS PAPUAN US F. Muell. in Austral. Journ. Pharm. 1 (1886) 
123; Bot. Centralbl. 27 (1886) 21? 
Lingoum saxatile Rumph. Herb. Amb. 2: 210. 

Lingoum saxatile Rumph., apparently a Pterocarpus, should 

* Retained name, Vienna Code; Amerimnon P. Br. (1756) is older. 

f See Prain, D., in Ann. Bot. Gard. Calcutta 10 1 (1904) 34, t. 8, who cites 
specimens from Amboina and who also cites Lacca lignum Rumph. as a 
synonym of Dalbergia parviflora Roxb. 



LEGUMINOSAE 



271 



be a very characteristic species, judging from the description of 
the fruits, which are stated to be twice as large as those of 
Pterocarpus indicus Willd., "quatuor vel quinque digitos transver- 
sales lati." Hasskarl, Neue Schlussel (1866) 41, has suggested 
that it may be Pterocarpus obtusatus Miq., which after all may 
be the correct disposition of it. Miquel's description is so very 
imperfect that no definite conception of his species can be gained 
from the description alone. I suggest that it may be F. Mueller's 
species, which is supposed to have large fruits. Pterocarpus 
blancoi Merr. should also be very closely allied. Dr. Alfred J. 
Ewart has kindly sent me a leaflet from the type of Pterocarpus 
papuanus F.-Muell., which is preserved in the national herbarium 
at Melbourne. The leaflet very closely resembles those of 
Pterocarpus indicus Willd. Doctor Ewart states that there are 
no fruits with the specimen, and the size of the fruits is not 
indicated by Mueller in the original description of the species. 

PTEROCARPUS SANTALI N US Linn. f. Suppl. (1881) 318. 
Sandalum rubrum Rumph. Herb. Amb. 2: 47. 

This is undoubtedly the correct disposition of Sandalum rub- 
rum, as suggested by Hasskarl, Neue Schlussel (1866) 47. It 
was not from Amboina. 

PONGAMIA * Ventenat 

PON GAM I A PIN NAT A (Linn.) comb. nov. 

Cytisus pinnatus Linn. Sp. PI. (1753) 741. 
Robinia mitis Linn. Sp. PI. ed. 2 (1763) 1044. 

Galedupa indica Lam. Encycl. 2 (1786) 594, excl. syn. Caju galedupa 
Rumph. 

Dalbergia arbor ea Willd. Sp. PI. 3 (1803) 901. 
Pongamia glabra Vent. Jard. Malm. 1 (1803) 28, t. 28. 
Galedupa pinnata Taub. in Engl. & Prantl Nat. Pflanzenfam. 3 3 (1891) 
344. 

Caju pinnatum 0. Kuntze Rev. Gen. PI. 1 (1891) 167. 
Pongamia mitis Merr. in Philip. Journ. Sci. 5 (1910) Bot. 101. 
Malaparius Rumph. Herb. Amb. 3: 183, t. 117. 
Malaparius e Nussanive Rumph. Herb. Amb. 5: 184. 

Amboina, Eri and Amahoesoe, Robinson PI. Rumph. Amb. 24-3, August 
and September, 1913, along the seashore, with normal fruits and with 
galls (Malaparius e Nussanive Rumph.!). 

Desrouss, in Lamarck Encycl. 3 (1791) 689, thought that Mala- 
parius might be near Pterocarpus ; and Loureiro, Fl. Cochinch. 
(1790) 431, erroneously cites it under his Pterocarpus fiavus, 
a species of doubtful status, this reduction being followed, how- 



* Retained name, Vienna Code; Galedupa Lam. (1786) is older. 



272 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ever, by Poiret, de Candolle, Don, Dietrich, Miquel, and a few 
other authors. Malaparius, however, is no Pterocarpus, but is 
manifestly identical with the plant commonly called Pongamia 
glabra Vent. The Rumphian figure is an excellent one, and the 
description applies perfectly except as to the statement that 
the fxowers are yellow; this may have been due to a mixture of 
material, or Rumphius may have had old flowers. The flowers 
are usually white or pink, turning somewhat yellowish in age. 
Prain, Journ. As. Soc. Beng. 68 2 (1897) 95, seems to have been 
the first author correctly to reduce Malaparius to Pongamia. 
Malaparius flavus Miq., Fl. Ind. Bat. 1 (1858) 1082, based on 
Sumatran specimens, the generic name from Rumphius, is ap- 
parently the form of Pongamia pinnata (Linn.) Merr. described 
by Hasskarl as Pongamia xerocarpa (P. glabra Vent. var. xero- 
carpa Prain) . 

I have taken up the earliest Linnean specific name for this 
widely distributed species, as I am now convinced that the speci- 
men in the Linnean herbarium, which is the common Pongamia 
glabra Vent., is the actual type of Cytisus pinnatus Linn., as 
indicated by the fact that Linnaeus not only gives a biblio- 
graphical reference to Plukenet, but also adds a short descrip- 
tion manifestly taken from an actual specimen. 

DERRIS * Loureiro 

DERRIS TR1FOLIATA Lour. Fl. Cochinch. (1790) 433. 

Robinia uliginosa Roxb. ex Willd. Sp. PL 3 (1800) 1133. 
Dalbergia heterophylla Willd. 1. c. 901. 
Pongamia uliginosa DC. Prodr. 2 (1825) 416. 

Derris forsteniana Blume ex Miq. Fl. Ind. Bat. 1 1 (1855) 144, t. 3. 
Tuba siliquosa Rumph. Herb. Amb. 5: 41, t. 25, f. 2. 
Amboina, near the town of Amboina, Robinson PI. Rumph. Amb. 537, 
October 25, 1913, along tidal streams, locally known as daun talahohor. 

Rumphius's figure is an excellent one of the widely distributed 
form commonly known as Derris uliginosa Roxb., but for which 
I maintain the earlier name, Derris trifoliata Lour. Loureiro's 
type is not extant, at least no specimen of his species is among 
his plants in the herbarium of the British Museum. Dalbergia 
heterophylla Willd. is exactly the same as Derris uliginosa Roxb. 
and has priority over Roxburgh's name in case Derris trifoliata 
Lour, be abandoned. Hasskarl, Neue Sehlussel (1866) 93, re- 
duced Tuba siliquosa Rumph. to Derris forsteniana Blume, which 
is a synonym of the above species. 

* Retained name, Vienna Code; Salken Adans. and Solori Adans. (1783), 
Deguelia Aubl. (1775), and Cylizoma Neck. (1790) are older. 



LEGUMINOSAE 



273 



DERRIS HEPTAPHYLLA (Linn.) comb. nov. 

Sophora heptaphylla Linn. Sp. PL (1753) 373, excl. syn. Plukenet. 
Pongamia sinuata Wall. Cat. (1832) no. 5911, nomen nudum. 
Derris sinuata Benth. ex Thw. Enum. PI. Zeyl. (1859-64) 93. 
Pterocarpus diadelphus Blanco Fl. Filip. (1837) 563. 
Derris diadelpha Merr. in Philip. Journ. Sci. 5 (1910) Bot. 103. 
Funis convolutus Rumph. Herb. Amb. 5: 69, t. 37, f. 1. 

This species is not represented in our Amboina collections, 
but unquestionably Funis convolutus Rumph. is identical with 
the widely distributed Indo-Malayan form commonly known as 
Derris sinuata Benth. Hasskarl, Neue Schliissel (1866) 96, after 
Miquel, has suggested that it may be the same as Derris montana 
Benth., which is hardly possible in view of the characters and 
geographic distribution of that species. The type of Sophora 
heptaphylla Linn, is Fl. Zeyl. 10 U, and Hermann's specimen on 
which Fl. Zeyl. 10U was based is Derris sinuata Benth. The 
reference to Plukenet, included by Linnaeus in the original de- 
scription as a doubtful synonym, cannot possibly be interpreted 
as the type. 

DERRIS ELLIPTICA (Roxb.) Benth. in Journ. Linn. Soc. Bot. 4 (1860) 
Suppl. 111. 

Galedupa elliptica Roxb. Hort. Beng. (1814) 53, nomen nudum, Fl. 

Ind. ed. 2, 3 (1832) 242. 
Tuba radicum alba Rumph. Herb. Amb. 5: 37, t. 23. 

This is not represented in our Amboina collections. The 
figure is unmistakably that of a species of Derris, either identical 
with, or very closely allied to, Derris elliptica Benth. The in- 
dicated use of the plant, for poisoning fish, is also a Derris 
character, several of the Malayan species being thus used, in- 
cluding Derris elliptica Benth. Hasskarl, Neue Schliissel (1866) 
92, considers the Rumphian figure and description to be referable 
to Millettia sericea W. & A., but this reduction is certainly 
incorrect. 

Perhaps referable to the same species of Derris as the above 
is Tuba radicum nigra Rumph. Herb. Amb. 5 : 38. It is used for 
the same purposes as Tuba radicum alba, and is at least a species 
of Derris. Hasskarl, Neue Schliissel (1866) 92, merely in- 
dicated that it belongs in the Dalbergiae. 

INOCARPUS Forster 
INOCARPUS EDULIS Forst. Char. Gen. (1776) 66, t. 33. 
Bocoa edulis Baill. Adansonia 9 (1868-70) 237. 
Gajanus edulis O. Kuntze Rev. Gen. PL 1 (1891) 189. 
Gajanus Rumph. Herb. Amb. 1: 170 t. 65. 

Amboina, Hitoe lama, Robinson PI. Rumph. Amb. 5U0, November, 1913, 

144971 18 



274 



RUMPHIUS'S HERBARIUM AMBOINENSE 



in light forests, altitude 50 meters; Amboina town, Robinson PL Rumph. 
Amb. 541, October 25, 1913, from cultivated trees, locally known as guyang 
and daun gayang. 

This reduction of Gajanus seems first to have been suggested 
by Lamarck, Encycl. 3 (1789) 253, which has been accepted by 
all succeeding authors and is the correct disposition of it. 

A BR US Linnaeus 

ABRUS PRECATORIUS Linn. Syst. ed. 12 (1767) 472. 

Glycine abrus Linn. Sp. PI. (1753) 753. 
Abrus frutex Rumph. Herb. Amb. 5: 57, t. 32. 

Amboina, Binting, Robinson PI. Rumph. Amb. 527, September, 1913, 
locally known as saga, saga alus, and daun saga. 

This species is too well known to need discussion. The orig- 
inal reduction of Abrus frutex to Glycine abrus Linn.= Abrus 
precatorius Linn, was made by Linnaeus, in Stickman Herb. 
Amb. (1754) 19, Amoen. Acad. 4 (1759) 128, Syst. ed. 10 (1759) 
1173, Sp. PL ed. 2 (1763) 1025. 

CLITORIA Linnaeus 

CLITORIA TERNATEA Linn. Sp. PI. (1753) 753. 

Flos coeruleus Rumph. Herb. Amb. 5: 56, t. 31. 

Amboina, Gelala, Robinson PI. Rumph. Amb. 543, August 25, 1913, along 
roadsides at low altitudes, locally known as bunga sayor and sayor katjang. 

This species is too well known to need discussion. Flos coeru- 
leus was first reduced to Clitoria ternatea by Linnaeus, in Stick- 
man Herb. Amb. (1754) 19, and has been consistently cited here 
by all succeeding authors. 

GLYCINE Linnaeus 

GLYCINE MAX (Linn.) comb nov. 

Phaseolus max Linn. Sp. PI. (1753) 725. 
Dolichos soja Linn. Sp. PI. (1753) 727. 
Soja hispida Moench. Meth. (1794) 153. 

Glycine hispida Maxim, in Bull. Acad. Petersb. 18 (1873) 398. 
Glycine soja S. & Z. in Abh. Akad. Muench. 4 2 (1843) 119. 
Glycine ussuriensis Kegel & Maack Tent. PI. Ussur. (1861) 50. 
Soja max Piper in Journ. Am. Soc. Agron. 6 (1914) 84. 
Cadelium Rumph. Herb. Amb. 5: 388, t. 140. 

This species is not represented in our Amboina collections, but 
the Rumphian figure is an excellent representation of the widely 
cultivated and well-known soy bean. It was originally reduced 



LEGUMINOSAE 



275 



by Linnaeus to Phaseolus max, in Stickman Herb. Amb. (1754) 
23, Amoen. Acad. 4 (1759) 132, Syst. ed. 10 (1759) 1162, Sp. 
PL ed. 2 (1763) 1018, in which he was followed by Burman 
f., Willdenow, Persoon, Poiret, Don, and other authors. Lour- 
eiro, Fl. Cochinch. (1790) 441, correctly referred it to Dolichos 
soja Linn., which is a synonym of Phaseolus max Linn. =Glycine 
max (Linn.) Merr. By Henschel and by Pritzel it has been 
also correctly referred to Soja hispida Moench., another synonym 
of Glycine max Merr. Miquel, Fl. Ind. Bat. I 1 (1855) 197, 
erroneously referred it to Phaseolus radiatus Linn. 

Phaseolus max Linn, has been considered a true Phaseolus and 
a synonym of P. radiatus Linn, by nearly all recent authors. 
It is clearly the soy bean, identical with Glycine hispida Maxim., 
and the specific name max should be maintained for the soy 
bean, whether Glycine or Soja be recognized as its generic name. 
Piper has declared in favor of the genus Soja, chiefly for the 
reason that of the eight species originally described by Linnaeus 
in Glycine, but a single one, G. javanica Linn., now remains in 
the genus, the other seven having been removed by subsequent 
authors to Apios, Kraunhia, Abrus, Rhynchosia, Amphicarpaea, 
and Fagelia, respectively. However, I am content to determine 
the type of the genus Glycine by elimination, which well maintain 
Glycine in its generally accepted sense with G. javanica Linn, 
as its type. 

Prof. C. V. Piper has cleared up the synonymy of this com- 
monly cultivated species; and with the aid of extensive data, 
supplied by Sir David Prain, he has clearly shown that Phaseolus 
max Linn, is identical with the commonly cultivated and well- 
known soy bean.* 

ERYTHRINA Linnaeus 

ERYTHRINA FUSCA Lour. Fl. Cochinch. (1790) 427. 

Erythrina ovalifolia Roxb. Hort. Beng. (1814) 53, Fl. Ind. ed. 2, 3 
(1832) 254. 

Gelala aquatica Rumph. Herb. Amb. 2: 235, t. 78. 

This characteristic species is not represented in our Amboina 
collections. The reduction of Gelala aquatica was made by 
Loureiro, Fl. Cochinch. (1790) 427, in the original description 
of the species. Erythrina ovalifolia Roxb., the name commonly 
employed in current botanical literature, is certainly a synonym. 

* For a very full discussion of the case see Piper, C. V. The name of 
the soy bean; a chapter in its botanical history. Journ. Am. Soc. Agron. 
6 (1914) 75-84. 



276 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ERYTHRINA VARIEGATA Linn, in Stickman Herb. Amb. (1754) 10, 
Amoen. Acad. 4 (1759) 122 (type!). 
Erythrina picta Linn. Sp. PI. ed. 2 (1763) 993. 
Gelala alba Rumph. Herb. Amb. 2: 234, t. 77. 

This species is not represented in our Amboina collections. 
The form so excellently figured by Rumphius occurs in the Philip- 
pines, in Palawan and in Mindanao, and on Corregidor Island in 
cultivation and is in all respects, except in its variegated leaves, 
the same as the common and widely distributed plant commonly 
known as Erythrina indica Lam. The Rumphian figure and 
description are the whole basis of Erythrina variegata Linn, 
and in part the basis of E. picta Linn. Strictly, the specific 
name variegata should be adopted to include not only the 
form with the variegated leaves, but also the much commoner 
and widely distributed form with uniformly green leaves, E. 
indica Lam. The differences between the two are no greater, 
for the purpose of distinguishing species or varieties, than be- 
tween the various color forms of Graptophyllum pictum (Linn.) 
Griff, or of Codiaeum variegatum (Lour.) Blume. 

Var. ORI ENTALIS (Linn.) comb. nov. 

Erythrina corallodendron Linn. var. orientalis Linn. Sp. PL (1753) 
706. 

Tetradapa javanorum Osbeck Dagbok Ostind. Resa (1757) 93. 
Erythrina indica Lam. Encycl. 2 (1785) 391. 
Erythrina orientalis Murr. Comm. Gotting. 8 (1787) 35, t. 1. 
Erythrina lithosperma Blume Cat. Gew. Buitenz. (1823) 92. 
Gelala litorea Rumph. Herb. Amb. 2: 230, t. 76. 
Amboina, near the town of Amboina, Robinson PL Rumph. Amb. 533, 
August 8, 1913, along the seashore, locally known as daun gelala. 

Gelala litorea Rumph. was originally reduced by Linnaeus to 
Erythrina corallodendron Linn., in Stickman Herb. Amb. (1754) 
10, Amoen. Acad. 4 (1759) 122, Syst. ed. 10 (1759) 1155; 
and in the second edition of the Species Plantarum, (1763) 993, 
it was definitely reduced to the var. orientalis Linn. Erythrina 
corallodendron Linn, is a mixture of an American species and 
what is now generally known as Erythrina indica Lam., here 
called Erythrina variegata Linn. var. orientalis (Linn.) Merr. 
Erythrina corallodendron Linn, is apparently typified by the 
American plant. Tetradapa javanorum Osbeck, which does not 
appear in Index Kewensis, is manifestly the same as Erythrina 
indica Lam., the type being from western Java. It is suspected 
that the forms from Java and China, briefly mentioned by 
Rumphius, are referable here, but the data given are too in«- 
definite for their certain determination. 



LEGUMINOSAE 



277 



MUCUNA * Adanson 

MUCUNA GIGANTEA (Willd.) DC. Prodr. 2 (1825) 405. 

Dolichos giganteus Willd. Sp. PI. 3 (1800) 1041. 

Carpopogon giganteum Roxb. Hort. Beng. (1814) 54. 

Zoopthalmum giganteum Prain in Journ. As. Soc. Beng. 66 2 (1897) 68. 

Lobus litoralis Rumph. Herb. Amb. 5: 10, t. 6. 
Amboina, Wae, Robinson PI. Rumph. Amb. 567, November 26, 1913, along 
the seashore, locally known as bharu laut. 

This was erroneously reduced by Loureiro, Fl. Cochinch. 
(1790) 456, to Citta nigricans Lour. = Mucuna nigricans (Lour.) 
Steud. Loureiro's species was described from Cochin-China 
material, and is entirely different from the plant that Rumphius 
described and figured, belonging in the section Citta, the pods 
with prominent oblique plaits across their faces. Most authors 
who have had occasion to cite the Rumphian figure have followed 
Loureiro's erroneous reduction. The forms indicated by Rumph- 
ius, 1. c, as nigra and maculata are probably merely variants 
of this widely distributed species. 

MUCUNA PRURIENS (Linn.) DC. Prodr. 2 (1825) 405. 

Dolichos pruriens Linn, in Stickman Herb. Amb. (1754) 23, Amoen. 

Acad. 4 (1759) 132, Syst. ed. 10 (1759) 1162 (type!). 
Stizolobium pruriens Medic, in Vorles. Churpf. Phys. Ges. 2 (1797) 
399. 

Carpopogon pruriens Roxb. Hort. Beng. (1814) 54. 
Negretia pruriens Blanco Fl. Filip. ed. 2 (1845) 411. 
Cacara pruritus Rumph. Herb. Amb. 5: 393, t. H2. 

This species is not represented in our Amboina collections, 
but I have before me a single pod of the species originating 
in Amboina, received from the botanic garden, Buitenzorg, Java, 
through Prof. C. V. Piper, of the United States Department of 
Agriculture; it is apparently identical with the widely dis- 
tributed, low-altitude, Philippine form, but is not the same as 
much of the material in various herbaria labelled Mucuna pru- 
riens DC. The Rumphian figure and description are the whole 
basis of Dolichos pruriens Linn., and the species must ac- 
cordingly be interpreted from it. In the second edition of his 
Species Plantarum (1763) 1019, Linnaeus added certain ref- 
erences, to Jacquin, to Sloane, and perhaps the one to Rheede, that 
represent a species quite different from the Philippine and 
Moluccan Mucuna pruriens (Linn.) DC, and from these 
references the species has, by most authors, been erroneously 
interpreted. 

* Retained name, Vienna Code; Zoopthamnum P. Br. and Stizolobium P. 
Br. (1756) are older. 



278 



RUMPHIUS'S HERBARIUM AMBOINENSE 



MUCUNA MINiATA sp. nov. § Citta (?). 

Parrana miniata Rumph. Herb. Amb. 5: 10. 

Amboina, Paso and Roemah tiga, Robinson PL Rumph. Amb. 566 (type), 
in thickets and forests at low altitudes, locally known as tali or tali-tali. 

Frutex scandens usque ad 15 m altus, inflorescentiis exceptis 
glaber ; f oliolis firme chartaceis, in siccitate nigricantibus, nitidis, 
oblongis ad oblongo-ovatis, usque ad 14 cm longis, prominente 
subcaudato-acuminatis, lateralibus obliquis, basi rotundatis vel 
obscurissime subcordatis, nervis utrinque circiter 4; inflor- 
escentiis circiter 20 cm longis, minute adpresse cinereo- 
pubescentibus atque pilis paucis urentibus instructis, fasciculatis, 
caulifloris; floribus miniatis, 6 ad 7 cm longis, calycis dentibus 
prominentibus, inferioribus angustis, 5 ad 7 mm longis. 

A scandent shrub attaining a height of 15 m, glabrous except 
the inflorescence, deciduous. Branches slender, smooth, terete. 
Petioles 6 to 8 cm long. Leaflets firmly chartaceous, blackish 
when dry, shining, 10 to 14 cm long, 5 to 6 cm wide, oblong-ovate 
to oblong, the terminal one equilateral, lateral ones more or less 
oblique, apex rather prominently subcaudate-acuminate, acumen 
1 to 1.5 cm long, blunt, base rounded or shallowly and obscurely 
cordate; lateral nerves about 4 on each side of the midrib, pro- 
minent, curved-ascending, obscurely anastomosing, the rachis 
extended about 2.5 cm beyond the insertion of the lateral leaflets, 
the petiolules black, 4 to 5 mm long. Plant leafless at time of 
flowering, the racemes up to 20 cm in length, fascicled on nodules 
along the trunk or larger branches, appressed cinereous-pubes- 
cent with short hairs, with a few, longer, yellowish-brown, stiff, 
stinging hairs intermixed, the indumentum more prominent on 
the calyx and pedicels than on the rachis. Flowers vermilion, 
6 to 7 cm long, somewhat curved, their pedicels mostly in pairs, 
slender, about 2 cm long, spreading. Calyx cup-shaped, the tube 
about 8 mm long, the teeth rather prominent, the upper one 
stouter than the others, 3 to 4 mm long, the two lateral ones 
slenderly acuminate, about 3 mm long, the lower one linear, 5 
to 7 mm in length. Standard 3.5 cm long, about 2 cm wide, 
blunt; wings rather strongly falcate, acuminate, up to 6 cm 
long, about 12 mm wide. Keel slightly longer than the wings, 
somewhat rostrate, outer margins ciliate in the lower part. 
Ovary and style hirsute. Fruit unknown. 

This species is well characterized by its glabrous leaves, rather 
large, crimson flowers, and prominent, slender calyx teeth. It 
belongs with a group of species found in New Guinea, including 
Mucuna novo-guineensis Scheff., M. bennettii F. Muell., and M. 
kraetkei Warb. Scheffer's species has bright orange flowers, 



LEGUMINOSAE 



279 



but the other two have red flowers. Warburg's species is dis- 
tinguished by having short calyx teeth. Mueller's species differs 
from the Amboina form in its distinctly larger flowers and very 
much longer calyx teeth. I am under obligations to Prof. A. 
J. Ewart, curator of the national herbarium at Melbourne, for 
a flower of d'Albertis's specimen, the type of Mucuna bennettii 
F. Muell., for purposes of comparison, as well as for a copy of 
Mueller's original description of the species. 

MUCUNA ATERRIMA (Piper & Tracy) comb. nov. 

Stizolobium aterrimum Piper & Tracy in U. S. Dept. Agr. Bur. PI. 

Ind. Bull. 179 (1910) 18, t. U, f. B, t. 7. 
Cacara nigra Rumph. Herb. Amb. 5: 381, t. 138. 
Cacara pilosa Rumph. Herb. Amb. 5: 392. 

This species is not represented in our Amboina collections. 
However, Cacara nigra is manifestly a Mucuna and is certainly 
synonymous with the form recently described by Piper and Tracy 
as Stizolobium aterrimum, as indicated by the Rumphian de- 
scription, rather poor figure, and especially the data as to the 
pods "ad ventrum tribus protuberantibus costis notati," and 
the seeds "primo rubentia, dein fusca, ac tandem nigerrima, 
glabra & splendentia." Stizolobium aterrimum was described 
from cultivated specimens originating in Brazil, Australia, 
Cochin-China, Barbadoes, Mauritius, Java, and Ceylon. Lin- 
naeus erroneously reduced Cacara nigra to Phaseolus unguicu- 
latus Linn., in Stickman Herb. Amb. (1754) 23, Amoen. Acad. 
4 (1759) 132, Syst. ed. 10 (1759) 1162, and to Dolichos unguicu- 
latus Linn., in Sp. PI. ed. 2 (1763) 1019, with which species it 
has nothing in common. He was followed by Burman f., 
Lamarck, Loureiro, Willdenow, Persoon, Henschel, Pritzel, and 
Miquel. By other authors, Wight and Arnott, Prodr. (1834) 
255, it was reduced, with doubt, to Mucuna capitata DC, which 
it certainly is not; while Miquel, Fl. Ind. Bat. I 1 (1855) 228, 
thought that it might be referable to Dolichos dasijcarpus 
Miq.= Dysolobium dolichoides (Roxb.) Prain. Piper and 
Tracy,* discuss Cacara nigra under Stizolobium capitatum 
(Roxb.) 0. Kuntze= Mucuna capitata DC, but on account of 
the unsatisfactory figure given by Rumphius failed to recognize 
it as their Stizolobium aterrimum, while excluding it as a syn- 
onym of S. capitatum O. Ktze. In spite of the poor figure the 
description of the fruit ar>d flower characters applies very closely 
in all particulars to Mucuna aterrima (Piper & Tracy) Merr. 



* The Florida velvet bean and related plants. U. S. Dept. Agr. Bur. PL 
Ind. Bull. 179 (1910) 13. 



280 



RUMPHIUS'S HERBARIUM AMBOINENSE 



From the rather full data given by Rumphius I am unable to 
distinguish from this species the form described by Rumphius 
as Cacara pilosa. He contrasts it with Cacara nigra, but ap- 
parently considered it distinct from the latter chiefly because 
it was a wild, not a cultivated plant. Henschel, Vita Rumph. 
(1833) 181, erroneously cites the name Cacara pilosa as a syn- 
onym of Phaseolus pilosus Klein and erroneously cites t. 1U2 
(=Mucuna pruriens DC.) as representing Cacara pilosa. As 
noted by Hasskarl, Neue Schlussel (1866) 136, it is close to 
Mucuna utilis Wall., but that species has the pods covered with 
a tawny pubescence. It differs from Mucuna velutina Hassk. 
in its purple, not white flowers. The chief objection to the iden- 
tification of Cacara pilosa Rumph. with Cacara nigra Rumph. 
= Mucuna aterrima (Piper & Tracy) Merr. is that Rumphius 
describes the former as having more pubescent leaves than the 
latter. 

DIOCLEA Humbolt, Bonpland, and Kunth 

DIOCLEA REFLEXA Hook. f. Niger Flora (1849) 306. 
Pa man a rubra Rumph. Herb. Amb. 5: 9, t. 5. 
Amboina, Toelehoe, Paso, and Amboina, in thickets at low altitudes and 
along the beach, Robinson PL Rumph. Amb. 397, November, 1913. 

Parrana rubra Rumph., one of the "sea beans," is fairly well 
described but very poorly figured by Rumphius. It has remained 
doubtful up to the present time, but I feel confident that it is 
the widely distributed Dioclea reflexa Hook. f. in spite of the 
poor figure. Miquel thought that the figure represented a species 
of Mucuna, and Teysmann thought it was a species of Derris. 
The description of the seeds applies closely to those of Dioclea 
reflexa, but the figure does not show their peculiar hilum 
character. 

CAN AVALI A * de Candolle 

CAN AVALI A MICROCARPA (DC.) comb. nov. 

Lablab microcarpus DC. Prodr. 2 (1825) 402 (type!). 

Canavalia turgida Grah. in Wall. Cat. (1832) no. 5534. 

Cacara litorea Rumph. Herb. Amb. 5: 390, t. 1J/.1, f. 1. 
Amboina, Paso, Robinson PL Rumph. Amb. 562, October 29, 1913, in 
fruit, climbing in thickets back of the beach; town of Amboina, Robinson 
PL Rumph. Amb. 561, October 26, 1913, in flower, in thickets along streams, 
locally known as katjang hor. 

Linnaeus, in Stickman Herb. Amb. (1754) 23, Amoen. Acad. 
4 (1759) 132, reduced Cacara litorea, with doubt, to Dolichos 



* Retained name, Brussels Congress; Canavali Adans. (1763) and Cle- 
mentea Cav. (1804) are older. 



LEGUMINOSAE 



281 



lablab Linn., with which species it has nothing in common. 
Lablab microcarpus DC. apparently supplies the oldest valid 
specific name for the species, which is here accepted, although 
unfortunately not an appropriate specific name, as the fruits are 
by no means "small" in this species. 

The synonymy of the species is rather complicated, and has 
been discussed by Prain,* who concludes that the name Canavalia 
obtusifolia DC. is properly applicable to this form with the 
wide pods and prominent keels or ridges. Canavalia obtusifolia 
DC. was based on Dolichos obtusifolius Lam., an American plant. 
Dolichos rotundifolius Vahl was also described from American 
specimens, and I believe that this name, as well as Canavalia 
obtusifolia (Lam.) DC, is a synonym of Canavalia lineata 
(Thunb.) DC. Katu-tjandi Rheede, Hort. Malabar. 8: 83, t. US, 
included pods of both Canavalia lineata DC. and C. turgida 
Gr ah. = Canavalia microcarpa (DC.) Merr. Hasskarl, Neue 
Schliissel (1866) 136, leaves Cacara litorea as a synonym of 
Lablab vulgaris Savi, which is an entirely wrong disposition of 
it. The Rumphian figure is very greatly reduced, whence de 
Candolle's inappropriate specific name. 

CANAVALIA LINEATA (Thunb.) DC. Prodr. 2 (1825) 404. 
Dolichos lineatus Thunb. Ft Jap. (1784) 280. 
Dolichos obtusifolius Lam. Encycl. 2 (1786) 295. 
Dolichos rotundifolius Vahl Symb. 2 (1791) 81. 
Canavalia obtusifolia DC. Prodr. 2 (1825) 404. 

Cacara litorea Rumph. Herb. Amb. 5: 390 quoad descr. p. p., non 
t. HI, t. 1. [See Canavalia microcarpa (DC.) Merr.]. 
Amboina, Hatiwe, Robinson PI. Rumph. Amb. 553, September 4, 1913, 
along the strand. 

This strand form, apparently always growing on the loose 
sand of the beach, is not to be confused with Canavalia micro- 
carpa (DC.) Merr. (C. turgida Grah.), which grows in thickets 
back of the beach. It is apparently included in the description 
of Cacara litorea Rumph., but is not the form figured by him; 
see Canavalia microcarpa (DC.) Merr., page 280. 

CANAVALIA GLADIATA (Jacq.) DC. Prodr. 2 (1825) 404. 
Dolichos gladiatus Jacq. Coll. 2 (1788) 276. 

Canavalia gladiata DC. var. machaeroides DC. Prodr. 2 (1825) 404 
(type!). 

Canavalia machaeroides DC. ex Steud. Nomencl. ed. 2, 1 (1840) 273 
(type!). 

Lobus machaeroides Rumph. Herb. Amb. 5: 376, t. 135, f. 1. 

This cultivated bean is not represented in our Amboina col- 



*Journ. As. Soc. Beng. 66 s (1897) 419. 



282 



RUMPHIUS'S HERBARIUM AMBOINENSE 



lections. The figure and description manifestly apply to the 
form generally known as Canavalia gladiata DC, which is found 
in scattered cultivation in most tropical countries. Linnaeus 
originally, but erroneously, reduced Lobus machaeroides to Doli- 
chos ensiformis Linn. = Canavalia ensiformis DC, in Stickman 
Herb. Amb. (1754) 23, Amoen. Acad. 4 (1759) 132, Syst. ed. 
10 (1759) 1162, Sp. PL ed. 2 (1763) 1022, in which he was 
followed by Burman f., Lamarck, and Loureiro. Willdenow, 
Sp. PL 3 (1800) 1039, referred it to Dolichos gladiatus, which 
as Canavalia gladiata (Jacq.) DC. is the correct disposition of 
it. Canavalia machaeroides DC. (C. gladiata DC. var. machaer- 
oides DC.) is based wholly on Rumphius and thus becomes a 
synonym of Canavalia gladiata (Jacq.) DC, although placed in 
Index Kewensis as a synonym of Canavalia cathartica Thouars. 

. PUERARIA de Candolle 

PUERARIA PHASEOLOIDES (Roxb.) Benth. in Journ. Linn. Soc. Bot. 9 

(1867) 125. 

Dolichos phaseoloides Roxb. Fl. Ind. ed. 2, 3 (1832) 316. 
Phaseolus minimus silvestris Rumph. Herb. Amb. 5: 387? 
Amboina, Soja road, Batoe gadjah, and Amboina (town), Robinson PL 
Rumph. Amb. 607, July and August, 1913, in thickets and waste places, 
sea level to an altitude of 200 meters, locally known as katjang panjang. 

The species briefly described by Rumphius is referred here 
with some doubt, although the description applies closely to 
Pueraria phaseoloides Benth. Hasskarl, Neue Schlussel (1866) 
136, thought that it might be Glycine mollis W. & A. Phaseolus 
minimus Rumph., described in the same chapter and figured, 
is Phaseolus aureus Roxb. 

CAJANUS * de Candolle 

CAJANUS CAJAN (Linn.) Millsp. in Field. Columb. Mus. Bot. 2 (1900) 

53. 

Cytisus cajan Linn. Sp. PL (1753) 739. 

Cytisus pseudo-cajan Jacq. Hort. Vind. 2 (1772) 54, t. 119. 
Cajan inodorum Medic, in Vorles. Churpf. Phys. Ges. 2 (1787) 363. 
Cajanus bicolor DC. Cat. Hort. Monsp. (1813) 85. 
Cajanus indicus Spreng. Syst. 3 (1826) 248. 
Phaseolus balicus Rumph. Herb. Amb. 5 : 377, t. 135, f. 2. 
Amboina, Koesoe koesoe sereh, Robinson PI. Rumph. Amb. 551, August 23, 
1913, locally known as kajan kay. 

Phaseolus balicus was originally reduced by Linnaeus to his 
Cytisus cajan, in Stickman Herb. Amb. (1754) 23, Amoen. Acad. 
4 (1759) 132, to which it certainly belongs. The oldest valid 



* Retained name, Brussels Congress; Cajan Adans. (1763) is older. 



LEGUMINOSAE 



283 



specific name is that supplied by the Linnean binomial, which is 
here accepted in place of the almost universally used Cajanus 
indicus Spreng. 

PHASEOLUS Linnaeus 

PHASEOLUS VULGARIS Linn. Sp. PL (1753) 723. 
Phaseolus scriptus Rumph. Herb. Amb. 5: 382? 
Faba rubra Rumph. Herb. Amb. 5: 382? 

Both of the above plants, very briefly described by Rumphius, 
were exotics, which had been introduced into Amboina and 
cultivated. Both of them may possibly be forms of Phaseolus 
vulgaris Linn., but this disposition of them is a mere guess. 

PHASEOLUS AUREUS Roxb. Hort. Beng. (1814) 55, nomen nudum, Fl. 
Ind. ed. 2, 3 (1832) 297. 
Phaseolus mungo auctt., non Linn. 
Phaseolus radiatus auctt., non Linn. 

Phaseolus minimus Rumph. Herb. Amb. 5: 386, t. 139, f. 2. 

The commonly cultivated and well-known mung bean is not 
represented in our Amboina collections. Linnaeus originally 
reduced Phaseolus minimus Rumph. to P. radiatus Linn., in 
Stickman Herb. Amb. (1754) 23, Amoen. Acad. 4 (1759) 132, 
Syst. ed. 10 (1759) 1162, in which he has been followed by all 
authors who have had occasion to cite the Rumphian figure and 
description. However, Phaseolus radiatus Linn, is not the mung 
bean, but is the form described by Roxburgh as Phaseolus sub- 
lobatus Roxb. Phaseolus max Linn., also referred to the mung 
by some authors, is the soy bean, Glycine max (Linn.) Merr. 
(Soja max Piper, Glycine hispida Maxim.) ; see page 274. 
Phaseolus mungo Linn, is a species distinct from P. aureus 
Roxb. (P. radiatus auctt.), the urd, or black gram, of India.* 

PHASEOLUS CALCARATUS Roxb. Hort. Beng. (1814) 54, nomen nudum, 
Fl. Ind. ed. 2, 3 (1832) 289. 
Phaseolus cylindraceus Rumph. Herb. Amb. 5: 389 ? 
Amboina, Kati-kati, Robinson PL Rumph. Amb. 233, October 6, 1913, 
in thickets, altitude about 90 meters. 

I am not sure that the plant here cited represents either 
Phaseolus cylindraceus Rumph. or P. calcaratus Roxb. The 
specimen is identical with the widely distributed, wild Philippine 
form that I have referred to Phaseolus calcaratus Roxb.f, which 

* Piper, C. V. Five Oriental species of beans. U. S. Dept. Agr. Bull. 
119 (1914) 1-32, t. 1-6, gives critical consideration of the mung bean and 
its allies. 

f Philip. Journ. Sci. 5 (1910) Bot. 132. 



284 



RUMPHIUS'S HERBARIUM AMBOINENSE 



differs from typical P. calcaratus in its more scandent habit 
and in its smaller seeds. 

VIGNA Savi 

VIGNA SINENSIS (Linn.) EndL ex Hassk. PI. Jav. Rar. (1848) 386. 

Dolichos sinensis Linn. Cent. PI. 2 (1756) 28, Amoen. Acad. 4 (1759) 

132, 326 (type!). 
Dolichos sinensis Rumph. Herb. Amb. 5 : 375, t. 1SU. 

This cultivated species is not represented in our Amboina col- 
lections, but the identity of Dolichos sinensis is unmistakable 
from Rumphius's figure and description. It is the commonly 
cultivated bean in southern China and the Indo-Malayan region 
with very long pods, indicated by Rumphius as "ulnam circiter 
longae." It is known in the Philippines as sitao, a name of 
Chinese origin, corresponding to the Chinese name tsjaitau quoted 
by Rumphius. Dolichos sinensis Rumph. is the whole basis 
of Dolichos sinensis Linn. Linnaeus took his name and brief 
description from Rumphius. I have not seen the original 
edition of the Centuria Plantarum 2 (1756), but in the reprint, 
Amoen. Acad. 4 (1759) 326, and in the reprint of Stickman's 
Herbarium Amboinense, 1. c. 132, the Rumphian reference is 
the whole basis of the species as proposed by Linnaeus. All 
early authors follow Linnaeus in quoting the Rumphian figure 
under Dolichos sinensis Linn., but some of the more recent ones, 
Hasskarl and Miquel, quote it under Vigna sinensis where it 
properly belongs. Vigna sinensis (Linn.) Endl., however, has 
not always been correctly interpreted by recent authors, but 
strictly must be limited to the form of the cow pea with the 
very long pods. 

VIGNA CYLINDRICA (Linn.) comb. nov. 

Phaseolus cylindricus Linn. Amoen. Acad. 4 (1759) 132 (type!). 
Dolichos catjang Linn. Mant. 2 (1771) 269. 
Phaseolus unguiculatus auctt., non Dolichos unguiculatus Linn. 
Phaseolus minor Rumph. Herb. Amb. 5: 383, t. 139, f. 1. 

This cultivated bean is not represented in our Amboina col- 
lections. The Rumphian figure and description of Phaseolus 
minor are the whole basis of Phaseolus cylindricus Linn. (1759), 
a name that has been overlooked by all subsequent authors, 
and which is not listed in Index Kewensis. It antedates 
Dolichos catjang Linn, and supplies the earliest valid specific 
name for the common cow pea. This form, with the short pods, 
by some authors has been considered either as identical with 
Vigna sinensis (Linn.) Endl. or as a variety of it. Modern 



LEGUMINOSAE 



285 



authors have generally considered it as a distinct species, some 
under the name Vigna unguiculata (Linn.) Walp. However, 
Dolichos unguiculatus Linn., the basis of Vigna unguiculata 
Walp., is Phaseolus unguiculatus (Linn.) Piper, in Torreya 12 
(1912) 190 (Phaseolus antillanus Urban), and has nothing to 
do with Vigna, with which it has been confused. Linnaeus 
cites the Rumphian name and figure in the original publication 
of Dolichos catjang (1771), in which he was followed by Bur- 
man f., Murray, Lamarck, Loureiro, Willdenow, Persoon, de 
Candolle, Don, and other authors. Rumphius described two 
forms, I albus and II ruber, which Hasskarl, Neue Schlussel 
(1866) 135, indicated as Vigna catjang var. alba Hassk. and 
V. catjang var. ruber Hassk., respectively. Both are probably 
merely cultural forms of the species. 

VIGNA MARINA (Burm.) comb. nov. 

Phaseolus marinus Burm. Index Universalis Herb. Amb. 7 (1755) [17] 
(type!). 

Dolichos luteus Sw. Prodr. Veg. Ind. Occ. (1788) 105. 
Vigna lutea A. Gray Bot. Wilkes U. S. Explor. Exped. (1854) 452. 
Phaseolus maritimus Rumph. Herb. Amb. 5: 391, t. 141, /. 2. 
Amboina, Hatiwe and Eri, Robinson PL Rumph. Amb. 536, September, 
1913, along the strand, locally known as katjang laut. 

The specific name above adopted for this well-known and widely 
distributed strand plant seems to be the oldest valid one for it. 
Burman's species is typified by the Rumphian figure and descrip- 
tion, the figure being an excellent representation of the plant 
commonly called Vigna lutea A. Gray. It is one of the few 
species published by Burman in the Index Universalis issued 
with Volume VII (Auctuarium) of the Herbarium Amboinense. 

PACHYRRHIZUS * Richard 

PACHYRRHIZUS EROSUS (Linn.) Urban Symb. Antil. 4 (1905) 311. 

Dolichos erosus Linn. Sp. PI. (1753) 726. 
Dolichos bulbosus Linn Sp. PI. ed. 2 (1763) 1021. 
Pachyrrhizus angulatus Rich, ex DC. Prodr. 2 (1825) 402. 
Pachyrrhizus bulbosus Kurz in Journ. As. Soc. Beng. 45 2 (1876) 246. 
Cacara erosa O. Kuntze Rev. Gen. PI. 1 (1891) 165. 
Cacara bulbosa Rumph. Herb. Amb. 5: 373, t. 132, f. 2. 

The common yam bean is not represented in our Amboina col- 
lections. It was introduced into Amboina from the Philippines, 
having been brought to the latter group from Mexico by the 
Spanish colonists. Cacara bulbosa was originally reduced by 

* Retained name, Vienna Code; Cacara Thou. (1805) is older. 



286 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Linnaeus to Dolichos erosus Linn., in Stickman Herb. Amb. 
(1754) 23, Amoen, Acad. 4 (1759) 132, Syst. ed. 10 (1759) 1163; 
but in the second edition of his Species Plantarum, ed. 2 (1763) 
1021, he placed it under Dolichos bulbosus Linn., which is 
manifestly a synonym of the earlier D. erosus Linn. By most 
recent authors it has been considered to be Pachyrrhizus angula- 
tus Rich., but this name must be abandoned for the much earlier 
Linnean one. 

PSOPHOCARPUS * Necker 

PSOPHOCARPUS TETRAGONOLOBUS (Linn.) DC. Prodr. 2 (1825) 403. 
Dolichos tetragonolobus Linn, in Stickman Herb. Amb. (1754) 23, 

Amoen. Acad. 4 (1759) 132, Syst. ed. 10 (1759) 1162, Sp. PI. ed. 

2 (1763) 1020 (type!). 
Botor tetragonoloba O. Kuntze Rev. Gen. PI. 1 (1891) 162. 
Lobus quadrangularis Rumph. Herb. Amb. 5: 374, t. 133. 

This well-known species is not represented in our Amboina 
collections, but is doubtless still cultivated there as it is in most 
parts of the Indo-Malayan region. The Rumphian figure and 
description are the whole basis of Dolichos tetragonolobus Linn., 
and it has been consistently cited by all authors under that 
name or its modern equivalent, Psophocarpus tetragonolobus DC. 

DOLICHOS Linnaeus 

DOLICHOS LABLAB Linn. Sp. PI. (1753) 725. 

Dolichos lignosus Linn. 1. c. 726. 

Dolichos albus Lour. PL Cochinch. (1790) 439. 

Lablab vulgaris Savi Diss. (1821) 19. 

Lablab perennans DC. Prodr. 2 (1825) 402. 

Cacara Rumph. Herb. Amb. 5: 378, t. 136. 

Cacara alba Rumph. Herb. Amb. 5: 380, t. 137. 
Amboina, Koesoekoesoe sereh, Robinson PI. Rumph. Amb. 552, August 23, 
1913, locally known as kakara puti. 

The specimen represents one of the several distinct forms of 
this variable species; in the shape of its pod it is somewhat 
different from both of the forms figured by Rumphius. Two 
distinct forms are figured and described by Rumphius: Cacara, 
with purple flowers, and Cacara alba, with white flowers. These 
are now, however, generally considered to represent a single 
species. Cacara was originally reduced by Linnaeus to Dolichos 
lignosus Linn., in Stickman Herb. Amb. (1754) 23, and has 
been cited by various authors under the names Dolichos altissi- 
mus Lour, and Lablab vulgaris Savi, both synonyms of Dolichos 



* Retained name, Vienna Code; Botor Adans. (1763) is older. 



OXALIDACEAE 



287 



lablab Linn. Cacara alba was cited by Loureiro in the original 
description of Dolichos albus Lour., Fl. Cochinch. (1790) 439, 
and has been cited by various authors under Lablab perennans 
DC, Lablab vulgaris Savi, and Lablab cultratus DC, all syn- 
onyms of Dolichos lablab Linn. 

Some authors, after Adanson, have maintained Lablab as a 
genus distinct from Dolichos Linnaeus, but I interpret Dolichos 
lablab Linn, as the type of the genus Dolichos, it being the first 
species cited in the Species Plantarum, while the generic descrip- 
tion, as given in the Genera Plantarum ed. 5 (1754) 324, conforms 
to the characters of this species. 

LEGUMI NOSiE OF UNCERTAIN STATUS 

Crotalaria montana VIII tsjeme tsjeme Rumph. Herb. Amb. 6: 146. 

A shrub from Macassar, Celebes, insufficiently described, but manifestly 
a representative of the Leguminosae. Hasskarl, Neue Schliissel (1866) 176, 
thought that it might be a species of Sophora, near S. glabra Hassk. 
Aeschynomene theophrasti Rumph. Herb. Amb. 5: 124. 

Burman f. referred this to Aeschynomene aspera Linn., but there is no 
warrant for this reduction. Rumphius merely discusses the plant as de- 
scribed by Theophrastus. 

OXALIDACEAE 

AVERRHOA Linnaeus 

AVERRHOA BILIMBI Linn. Sp. PL (1753) 428. 

Blimbingum teres Rumph. Herb. Amb. 1: 118, t. 36. 

Amboina, Binting, Robinson PL Rumph. Amb. 219, August 13, 1913, 
locally known as blimbing. 

This is one of the few Rumphian species that Linnaeus reduced 
in the first edition of his Species Plantarum, the Rumphian 
citation appearing in Sp. PL (1753) 428, and in all of Linnaeus's 
subsequent works in which the species is considered. The reduc- 
tion is certainly correct, and Linnaeus has been followed by all 
subsequent authors. 

AVERRHOA CARAMBOLA Linn. Sp. PI. (1753) 428. 

Prunum stellatum Rumph. Herb. Amb. 1: 115, t. 35. 
Amboina, from cultivated plants near the town of Amboina, Robinson 
PL Rumph. Amb. 218, August 23, 1913, locally known as blimbing manis. 

This Rumphian species, like the preceding one, was reduced 
by Linnaeus in the first edition of his Species Plantarum (1753) 
428, appears in the subsequent writings of the same author in 
which Averrhoa is considered, and the reduction, certainly cor- 
rect, has been accepted by all subsequent authors. 



288 



RUMPHIUS'S HERBARIUM AMBOINENSE 



BIOPHYTUM de Candolle 

BIOPHYTUM SENSITIVUM (Linn.) DC. Prodr. 1 (1824) 690. 
Oxalis sensitiva Linn. Sp. PL (1753) 434. 
Herba sentiens Rumph. Herb. Amb. 5: 301, t. 10U, f. 2. 
Amboina, Batoe mera, Robinson PL Rumph. Amb. 217, July 20, 1913, 
along ditches in rocky soil, altitude 5 to 10 meters. 

Herba sentiens Rumph. was reduced by Linnaeus to Oxalis 
sensitiva, in Stickman Herb. Amb. (1754) 22, Amoen. Acad. 4 
(1759) 131, Syst. ed. 10 (1759) 1038, Sp. PI. ed. 2 (1762) 622, 
which is certainly correct. The plant is now known as Biophy- 
tum sensitivum (Linn.) DC, and most authors who have cited 
the Rumphian figure since 1824 have so placed it. Miquel, 
however, Fl. Ind. Bat. I 2 (1858) 134, considers that the 
Rumphian figure represents Oxalis reinwardtii Zucc, which he 
maintains as a species distinct from Biophytum sensitivum 
(Linn.) DC. 

OXALIS Linnaeus 

OXALIS REPENS Thunb. Diss. Oxal. (1781) 16. 
Oxys lutea indica Rumph. Herb. Amb. 5: 277. 
Amboina, on walls in the town of Amboina, Robinson PI. Rumph. Amb. 
216, July 22, 1913, locally known as daun kelauwar. 

Burman f., Fl. Ind. (1768) 107, and Loureiro, Fl. Cochinch. 
(1790) 285, place Oxys lutea indica under Oxalis corniculata 
Linn., and most authors have considered Oxalis repens Thunb. 
to be a synonym of the older Linnean species. However, Dr. 
B. L. Robinson * has shown that two species are involved in 
what is generally called Oxalis corniculata Linn. The actual 
Amboina specimens, like most or all the material from the 
Malayan region usually called Oxalis corniculata Linn., are 
actually referable to Thunberg's species, as the two forms are 
distinguished by Doctor Robinson. 

RUTACEAE 

FAGARA Linnaeus 

FAGARA TORVA (F. Muell.) Engl, in Engl. & Prantl Nat. Pflanzenfam. 

3* (1895) 119. 
Zanthoxylum torvum F. Muell. Fragm. 7 (1871) 140. 
Zanthoxylum glandulosum T. & B. Cat. Hort. Bogor. (1866) 234, 

nomen nudum. 
Nugae sylvarum silvestris Rumph. Herb. Amb. 5: 124. 
Amboina, Paso, on trees at low altitudes, Robinson PI. Rumph. Amb. 250, 
October 29, 1913. 



* Journ. Bot. 44 (1906) 391. 



RUTACEAE 289 

Hasskarl, Neue Schliissel (1866) 99, has suggested that Nugae 
sylvarum silvestris might be a Zanthoxylum. With Amboina 
material that certainly represents the Rumphian species, this 
supposition can now be verified. The specimen is a very close 
match for Hochreutiner's PL Bogor. Exsiccatae No. 28, which 
is typical Zanthoxylum glandulosum T. & B., and which Hochreu- 
tiner interprets as typical Zanthoxylum torvum F. Muell. The 
same form is found in Luzon, Leyte, and Mindanao in the Philip- 
pines, so that the known range of the species is now from Luzon 
to Java, the Moluccas, and tropical Australia. 

FAGARA sp. 

Panax ? anisum DC. Prodr. 4 (1830) 254 (type!). 

Nothopanax ? anisum Miq. Fl. Ind, Bat. 1 1 (1857) 766 (type!). 

Anisum moluccanum Rumph. Herb. Amb. 2: 131, t. 1+2. 

This species is not represented in our Amboina collections. 
The description in all respects applies to Fagara (Zanthoxylum) , 
but without specimens I am unable definitely to refer Anisum 
moluccanum Rumph. to any described Malayan species. So far 
as can be determined from the description, Panax anisum DC. 
is based wholly on Rumphius, Nothopanax anisum Miq. being 
merely a transfer of de Candolle's name. Henschel erroneously 
referred the Rumphian species to Zanthoxylum aromaticum 
Willd., an American species. The figure closely resembles 
Fagara avicennae Lam. 

EVODIA Forster 

EVODIA LATI FOLIA DC. Prodr. 1 (1824) 724 (type!). 
Ampacus latifolia Rumph. Herb. Amb. 2: 186, t. 61. 

This species was based wholly on Rumphius and must be 
interpreted entirely from the Rumphian figure and description. 
Evodia latifolia DC. has been interpreted by recent authors 
as being represented by Philippine specimens collected by 
Cuming, but the Philippine material has been described under 
two different names, Evodia bintoco Blanco and Evodia philip- 
pinensis Merr. Burman f. thought that the plant figured by 
Rumphius might be a species of Rhus, Lamarck a Premma 
or Vitex, and Poiret, with doubt, an Aubertia. De Candolle, 
however, based his Evodia latifolia wholly on Rumphius. 
It has also been called Zanthoxylum latifolium Don, Fagara 
latifolia Roxb., and Zanthoxylum rumphianum Cham. It 
is undoubtedly a true Evodia. Miquel, Ann. Mus. Bot. Lugd. 
Bat. 3 (1867) 244, has redescribed Evodia latifolia DC. from 
Halmaheira specimens, and his description conforms closely 

144971 19 



290 



RUMPHIUS'S HERBARIUM AMBOINENSE 



to Philippine material referred to de Candolle's species. Am- 
boina specimens are desirable in order definitely to determine 
the true status of the species. 

EVODIA AMBOINENSIS sp. nov. 

Ampacus angustifolia Rumph. Herb. Amb. 2: 188, t. 62. 
Amboina, Koesoejoesoe sereh, Robinson PI. Rumph. Amb. 251 (type), 
October 3, 1913, in light woods at an altitude of about 225 meters; hills 
behind the town of Amboina, in light forests, Robinson PI. Rumph. Amb. 
252, October 27, 1913, locally known as gendarussa. 

Arbor parva, 8 ad 10 m alta, glabra, vel ramulis junioribus 
inflorescentiisque plus minusve cinereo-puberulis ; f oliis 3-f oliola- 
tis, foliolis petiolatis, firme chartaceis ad subcoriaceis, oblongis, 
usque ad 16 cm longis, acuminatis, basi acutis ad rotundatis, 
nervis utrinque 10 ad 12, subtus prominentibus ; paniculis axilla- 
ribus, pyramidatis, usque ad 13 cm longis, ramis patulis, mul- 
tifloris, floribus circiter 2.5 mm longis, ovario pubescente. 

A tree 8 to 10 m high, nearly glabrous, or the young branchlets 
and parts of the inflorescence more or less cinereous-puberulent. 
Branches terete, reddish-brown, often somewhat compressed at 
the nodes, smooth. Leaves opposite, 3-foliolate, their petioles 
4 to 7 cm long; leaflets in general oblong, firmly chartaceous to 
subcoriaceous, 9 to 16 cm long, 4 to 7 cm wide, entire, apex rather 
prominently acuminate, base acute to rounded, somewhat shining 
when dry, subolivaceous or somewhat pale, quite glabrous on 
both surfaces ; lateral nerves 10 to 12 on each side of the midrib, 
slender but prominent, anastomosing, the reticulations lax; pe- 
tiolules 3 to 5 mm long. Panicles axillary, pyramidal, slightly 
puberulent or nearly glabrous, up to 13 cm long, the branches 
spreading, the lower ones up to 7 cm long. Flowers white, 
numerous, their pedicels about 2 mm long. Sepals 4, ovate, 
obtuse, 0.5 mm long. Petals 4, elliptic-ovate, glabrous, slightly 
apiculate at the apex, 2 to 2.5 mm long; filaments elongated, 
glabrous; anthers 1.2 mm long. Ovary 4-lobed, pubescent. 
Fruit small, composed of two or three, nearly free, dehiscent 
cocci about 3.5 mm in length, the seeds blue-black, shining, about 
2 mm in diameter. 

Ampacus angustifolius has been confused with Evodia tri- 
phylla DC. by practically all authors since the publication of 
the latter species in de Candolle's Prodromus 1 (1824) 724, 
which in turn was based on Fagara triphylla Lam., Encycl. 2 
(1788) 447. The type of Fagara triphylla was a Philippine 
specimen, and it is Melicope triphylla (Lam.) Merr., in Philip. 
Journ. Sci. 7 (1912) Bot. 375, where the complicated synonymy 



RUTACEAE 



291 



of the species is discussed. It is manifest, however, that most 
authors interpreted Evodia triphylla (Lam.) DC. not from the 
Philippine specimen actually described, but from the Rumphian 
figure, for Lamarck in the original description of Fagara tri- 
phylla adds a reference to Ampacus angustifolia Rumph. The 
names Fagara triphylla Lam., Evodia triphylla DC, and Zan- 
thoxylum triphyllum Don all refer to Melicope triphylla (Lam.) 
Merr. as synonyms and cannot be applied to the Amboina plant. 
It was erroneously reduced by Miquel, Fl. Ind. Bat. I 2 (1859) 
671, to Zanthoxylum zeylanicum DC. 

Evodia triphylla (Lam.) DC. has been given a range of from 
Tenasserim to Japan southward through Malaya to New Guinea, 
and to it have been referred specimens representing at least 
four distinct species in two different genera. As already 
pointed out by me, true Zanthoxylum triphyllum hscm.^ Evo- 
dia triphylla DC. is Melicope triphylla (Lam.) Merr., a species 
confined to the Philippines. Chinese material generally, but 
erroneously, referred to Evodia triphylla DC. is Evodia pteleae- 
folia (Champ.) Merr., in Philip. Journ. Sci. 7 (1912) Bot. 377; 
material from India, the Malay Peninsula, and the Sunda Islands 
for the most part is Evodia lunur-ankenda (Gaertn.) Merr., 1. c. 
378; while the actual Ampacus angustifolius Rumph. of Amboina 
represents still another species, Evodia amboinensis Merr. de- 
scribed above. This is known only from Amboina, but ma- 
terial from other parts of the Moluccas and from New Guinea, 
erroneously referred by other authors to Evodia triphylla DC, 
may prove to be identical with the Amboina species. 

FLINDERSIA R. Brown 

FLINDERSIA AMBOINENSIS Poir. in Lam. Encycl. Suppl. 4 (1816) 650 
(type!). 

Flindersia radulifera Spreng. Geschicht. Bot. 2 (1818) 76 (type!), 

ex Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 208. 
Arbor radulifera Rumph. Herb. Amb. 3: 201, t. 129. 

This species is not represented in our Amboina collections. 
Arbor radulifera Rumph. is the whole basis of Flindersia am- 
boinensis Poir. and of F. radulifera Spreng. The latter name 
does not appear in Index Kewensis. All descriptions of the 
species published, up to and including that of A. de Candolle,* 
have been based on the data given by Rumphius, no botanist 
having had specimens. It is cultivated in the botanic garden 
at Buitenzorg, Java, according to botanical specimens named 

* Meliaceae in DC. Monog. Phan. 1 (1878) 735. 



292 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Flindersia amboinensis Poir. from "III-A-7" and distributed 
from that institution. 

MURRAYA * Koenig 

MURRAYA PAN ICULATA (Linn.) Jack in Malay Miscel. 1 (1820) 31. 

Chalcas paniculata Linn. Mant. 1 (1767) 68. 

Chalcas camuneng Burm. f., Fl. Ind. (1768) 104. 

Murray a scandens Hassk. in Abh. Naturf. Gesellsch. Halle 9 (1866) 
233 (Neue Schliissel 91) (type!). 

Camunium japonense Rumph. Herb. Amb. 5: 29, t. 18, f. 2. 

Camunium javanicum Rumph. Herb. Amb. 5: 27. 

Camunium vulgare Rumph. Herb. Amb. 5: 26, t. 17. 
AMBOINA, Robinson PL Rumph. Amb. 2h9, September 13, 1913, from 
cultivated trees in the town of Amboina, locally known as kamuneng; exactly 
Camunium japonense Rumph. ! 

While the figures given by Rumphius represent two distinct 
forms, it is very doubtful whether or not two species are rep- 
resented. Specimens are found in herbaria that apparently 
present all intergradations between the form with small leaflets 
(Camunium japonense Rumph.) and the form with fewer and 
larger leaflets (Camunium vulgare Rumph.) . If but one species 
be represented, then the oldest valid name is Murraya paniculata 
(Linn.) Jack, and pending a critical revision of the genus it is 
probably best to consider both forms described by Rumphius as 
representing one variable species. 

Linnaeus, in Stickman Herb. Amb. (1754) 18, Amoen. Acad. 
4 (1759) 128, erroneously reduced Camunium japonense to 
Vitex pinnata Linn., the latter being given by some authors as 
a synonym of Aglaia odorata Lour. However, Vitex pinnata 
Linn, was based on Fl. Zeyi. 415, and Hermann's specimen is 
Vitex altissima Linn.f The reduction of Camunium japonense 
to Murraya exotica Linn, was made by Lamarck, Encycl. 4 
(1797) 382, in which he has been followed by numerous authors. 

Camunium vulgare Rumph. was placed by Burman f., Fl. Ind. 
(1768) 104, as a synonym of Chalcas camuneng, but Chalcas 
camuneng Burm. f. was primarily based on Javan specimens 
from cultivated plants. Linnaeus based his Chalcas paniculata 
on Burman's species and also cited Rumphius. Whether the 
form actually in Burman's hands was the one with small leaflets 
(typical Murraya exotica Linn.) or with fewer and larger 
leaflets is impossible to determine from the original descrip- 
tion, but it was probably the former. Murraya sumatrana 

* Retained name, Brussels Congress; Camunium Adans. (1763), Chalcas 
Linn. (1767), and Bergera Koenig (1771) are older, 
f See Trimen, Fl. Ceyl. 3 (1895) 358. 



RUTACEAE 



293 



Roxb. is apparently quite the same as Camunium vulgare 
Rumph. ; while Murray a scandens Hassk., which does not appear 
in Index Kewensis, must be typified wholly by the Rumphian 
figure and description. The forms mentioned by Hasskarl, Neue 
Schliissel (1866) 91, as Camunium javanicum and Camunium e 
Macassar are both probably referable to Murraya paniculata 
(Linn.) Jack as here interpreted. 

FERONIA Correa 

FERONIA LIMONIA (Linn.) Swingle in Journ. Wash. Acad. Sci. 4 (1914) 

328. 

Schinus limonia Linn. Sp. PL (1753) 389. 

Limonia acidissima Linn. Sp. PI. ed. 2 (1762) 554. 

Feronia elephantum Correa in Trans. Linn. Soc. 5 (1800) 225. 

Anisifolium Rumph. Herb. Amb. 2: 133, t. U3. 

Nothing resembling Anisifolium appears in our Amboina col- 
lections. Regarding the Rumphian figure, Swingle states that 
it is "the wood apple or a very closely allied species." Linnaeus 
made the reduction to his Schinus limonia, in Stickman Herb. 
Amb. (1759) 9, Amoen. Acad. 4 (1759) 121, Syst. ed. 10 (1759) 
1034, and in his Species Plantarum, ed. 2 (1762) 554, to Limonia 
acidissima, the latter reduction being followed by most authors. 
Hamilton placed it under Feronia elephantum Corr., while Hass- 
karl places it under Hesperethusa acidissima Roem., a synonym 
of Feronia limonia (Linn.) Swingle. Rumphius's material was 
from Java where the species is still cultivated. 

AEGLE * Correa 

AEGLE MARMELOS (Linn.) Correa in Trans. Linn. Soc. 5 (1800) 223. 
Crataeva marmelos Linn. Sp. PI. (1753) 444. 
Bilacus marmelos O. Kuntze Rev. Gen. PI. 1 (1891) 98. 
Bilacus Rumph. Herb. Amb. 1: 197, t. 81. 
Bilacus taurinus Rumph. Herb. Amb. 1: 199. 

This reduction of Bilacus was first made by Linnaeus, in 
Stickman Herb. Amb. (1754) 8, Amoen. Acad. 4 (1759) 120, 
Sp. PI. ed. 2 (1762) 637, and is certainly the correct disposition 
of it. It was described and figured from cultivated specimens. 

I am unable definitely to place Bilacus amboinensis silvestris 
Rumph., Herb. Amb. 1 : 200, t. 28. It was erroneously reduced 
by Linnaeus, in Stickman Herb. Amb. (1754) 8, Amoen. Acad. 
4 (1759) 120, to Crataeva tapia Linn. Hamilton suggested that 
it might be an undescribed species of Aegle, but this is improb- 
able. The figure strongly resembles Crataeva religiosa Forst., 



* Retained name, Brussels Congress; Belou Adans. (1763) is older. 



294 



RUMPHIUS'S HERBARIUM AMBOINENSE 



but the description does not conform to Crataeva in the seed 
characters : "In centro quatuor vel quinque locantur ossicula, a 
sese invicem remota, lanosa seu pilosa instar seminum Gossypii." 
In the original description Rumphius states that the flowers 
were unknown to him; but in the plate, drawn after he became 
blind, the flowers are shown. His original material was from 
Manipa, Sula Islands, and Celebes, but in the supplementary 
data taken from the Auctuarium and appended to the original 
description he cites specimens from Leytimor, Amboina, and 
states regarding the fruit : "Fructus est instar ovi minoris avis 
Casuarii." It is possible that the description was based on 
material from two different species. The figure, however, con- 
forms to the characters of Bilacus amboinensis silvestris as 
given in the text. In some respects the figure and description 
suggest Chaetospermum, typified by the Philippine Chaetosper- 
mum glutinosum (Blanco) Swingle, but Bilacus amboinensis 
silvestris Rumph. can hardly belong in the Rutaceae. 

FORTUNELLA Swingle 

FORTUNELLA J A PON I CA (Thunb.) Swingle in Journ. Wash. Acad. Sci. 
5 (1915) 171, /. 3. 
Citrus japonica Thunb. Nov. Act. Upsal. 3 (1780) 199. 
Citrus madurensis Lour. Fl. Cochinch. (1790) 467. 
Citrus inermis Roxb. Fl. Ind. ed. 2, 3 (1832) 393. 
Limonellus madurensis Rumph. Herb. Amb. 2: 110, t. SI. 

Loureiro, Fl. Cochinch. (1790) 467, first reduced Limonellus 
madurensis to his Citrus madurensis, which Swingle considers 
to be a synonym of Citrus japonica Thunb. =Fortunella japonica 
Swingle. The plant is a native of China, and the only possible 
objection I note to the reduction of Limonellus madurensis is 
that Citrus japonica is not known to occur in the Malay Archi- 
pelago. However, Rumphius states that the plant was cultivated 
in Madura and at Batavia, Java, where it may have been intro- 
duced from China and where it has failed to persist. Hasskarl, 
Neue Schlussel (1866) 32, suggests Atalantia monophylla DC. 
as the proper reduction of Limonellus madurensis Rumph., but 
judging from the figure and description given by Rumphius the 
plant can hardly have been an Atalantia. 

MEROPE M. Roemer 

MEROPE ANGULATA (Willd.) Swingle in Journ. Wash. Acad. Sci. 5 
(1915) 423. 

Citrus angulata Willd. Sp. PL (1800) 1426 (type!). 
Sclerostylis spinosa Blume Bijdr. (1825) 134. 
Limonia spinosa Spreng. Syst. Veg. 4 2 (1827) 162. 



RUTACEAE 



295 



Glycosmis spinosa Dietr. Syn. PI. 2 (1840) 1409. 
Merope spinosa M. Roem. Syn. Mon. Hesp. 1 (1846) 44. 
Limonia angulosa W. & A. ex Miq. Fl. Ind. Bat. 1 2 (1859) 521 
(type!). 

Atalantia longispina Kurz in Journ. As. Soc. Beng. 41 2 (1872) 295. 
Gonocitrus angulatus Kurz 1. c. 42 2 (1874) 228, t. 18. 
Paramignya longispina Hook. f. Fl. Brit. Ind. 1 (1875) 511. 
Paramignya angulata Kurz in Journ. As. Soc. Beng. 43 3 (1874) 135. 
Atalantia spinosa Hook. f. ex Koord. Exkurs. Fl. Java 2 (1912) 427. 
Limonellus angulosus Rumph. Herb. Amb. 2: 110, t. 32. 

The above formidable list of synonyms is copied from Swingle's 
paper on Merope* He has there given a critical consideration 
of Merope angulata (Willd.) Swingle and its numerous syn- 
onyms. Rumphius's figure and description are the whole basis 
of Citrus angulata Willd. and hence typify the species, which 
on Wight and Arnott's suggestion was transferred to Limonia as 
L. angulosus W. & A. by Miquel ; it is, therefore, also the type of 
Limonia angulosa W. & A. It is also the name-bringing syn- 
onym of both Gonocitrus angulatus Kurz and Paramignya angu- 
lata Kurz. Kurz was the first to recognize the identity of 
Limonellus angulosus Rumph. and its true relationships. 

CITRUS Linnaeus 

A number of representatives of the genus Citrus are figured 
and described by Rumphius, and these have been variously inter- 
preted by botanists. Some maintain that the species of this 
genus are reducible to a few polymorphous types with nu- 
merous varieties, while others maintain that the genus is 
composed of a large number of closely allied species. The 
probabilities are that there are relatively few species and that 
these have yielded numerous horticultural forms and hybrids ; 
it is not at all improbable that natural hybrids occur. In such 
a genus as Citrus it is naturally to be expected that authors 
have widely differed in interpreting the forms described by 
Rumphius. In relatively few cases is it possible definitely to 
determine from his figures and descriptions alone the exact 
status of the several forms in our present system of classifica- 
tion. The Robinson Amboina collection presents but two species 
of Citrus, the common lime and the common pomelo, and na- 
turally helps but little in determining the status of the Rumphian 
species or forms. Until extensive field work is prosecuted in 
the whole Malayan region and until the entire genus Citrus 
undergoes a very thorough and critical revision, any attempt 

* Merope angulata, a salt-tolerant plant related to Citrus, from the 
Malay Archipelago. Journ. Wash. Acad. Sci. 5 (1915) 420-425. 



296 



RUMPHIUS'S HERBARIUM AMBOINENSE 



to interpret the Rumphian figures and descriptions must be 
unsatisfactory. Linnaeus, in Stickman Herb. Amb. (1754) 9, 
Amoen. Acad. 4 (1759) 121, reduced most of the forms that 
Rumphius figured to Citrus medica Linn, and Citrus aurantium 
Linn., which disposition of them is not at all satisfactory; yet 
succeeding attempts to interpret them are hardly more satis- 
factory. It is probable that all or most of them have been de- 
scribed by succeeding authors, but from descriptions alone it 
is quite impossible to determine precisely to what forms or 
species the Rumphian plants should be reduced. 

CITRUS AURANTI FOLIA (Christm.) Swingle in Journ. Wash. Acad. Sci. 
3 (1913) 465. 

Limonia aurantifolia Christm. Pflanzensyst. 1 (1777) 618. 
Limonia acidissima Houtt. Handl. 2 2 (1774) 444, non Linn. 
Citrus lima Lunan Hort. Jamaic. 2 (1814) 451. 
Citrus acida Roxb. Fl. Ind. ed. 2, 3 (1832) 390. 
Citrus notissimus Blanco Fl. Filip. (1837) 607. 
Citrus limonellus Hassk. in Flora 25 (1842) Beibl. 43. 
Limonellus Rumph. Herb. Amb. 2: 107, t. 29. 
Amboina, Binting, Robinson PI. Rumph. Amb. 2U7 , August 13, 1913, along 
roadsides, locally known as limon china. 

This is the common lime. Limonellus was erroneously re- 
duced by Linnaeus, in Stickman Herb. Amb. (1754) 9, Amoen. 
Acad. 4 (1759) 121, to Citrus aurantium Linn. Both Houttuyn 
and Christmann cite the Rumphian figure, and Christmann's 
specific name is the oldest valid one for the species, as shown 
by Swingle. Blume cites the Rumphian plant as a synonym of 
Citrus javanica Blume, Bijdr. (1825) 140, while Hasskarl, Neue 
Schlussel (1866) 32, places it under Citrus limonellus Hassk., 
the specific name being apparently taken from Rumphius. The 
form figured on the same plate by Rumphius, fig. A> Hasskarl, 
1. c, considers as a variety, calling it Citrus limonellus var. 
oxycarpus Hassk. 

CITRUS MAXIMA (Burm.) comb. nov. 

Aurantium maximum Burm. ex Rumph. Herb. Amb. Auctuarium 

(1755) Ind. Univ. [16] (type!). 
Citrus grandis Osbeck Dagbok Ostind. Resa (1757) 98. 
Citrus aurantium Linn. var. grandis Linn. Sp. PI. (1753) 783. 
Citrus decumana Linn. Syst. ed. 12 (1767) 508. 
Li mo decuman us Rumph. Herb. Amb. 2: 96, t. 2U, /• 2. 
Amboina, Way tommo, Robinson PI. Rumph. Amb. 248, August 17, 1913, 
altitude about 60 meters, locally known as limon. 

The common pomelo is commonly known as Citrus decumana 
Linn., but there are at least two older names, the older of 
which is here accepted. Aurantium maximum Burm., a name 



RUTACEAE 



297 



not listed in Index Kewensis, is validly published by Burman in 
the general index to the Herbarium Amboinense by citation of 
the Rumphian name and description. It is to be noted that 
Burman on page [11] of the Index Universalis refers Limo 
decumanus Rumph. to Citrus aurantium Linn. Swingle, in 
Sargent PL Wils. 2 (1914), has already pointed out that Citrus 
grandis Osbeck antedates the publication of Citrus decumana 
Linn, and accordingly accepted Osbeck's specific name. Three 
forms described by Rumphius, //, III, and IV, are all probably 
referable here; although Hasskarl, Neue Schlussel (1866) 31, 
places them under the varieties pyriformis Hassk., leucosarca 
Hassk., and dulcis Hassk., respectively. 

CITRUS OBVERSA Hassk. Cat. Hort. Bogor. (1844) 218. 
Limo ferus Rumph. Herb. Amb. 2: 106, t. 26, f. 3, t. 28. 

In this reduction I merely follow Hasskarl, Neue Schlussel 
(1866) 32. The figures strongly resemble forms of Citrus 
hystrix DC, as de Candolle's species is currently interpreted; 
and Limo ferus Rumph., with Citrus obversus Hassk., may be 
merely a form of this polymorphous species. At any rate, 
the specific status of Citrus obversa Hassk. is very doubtful. 

CITRUS BERGAM I A Risso Hist. Nat. Or. Europ. Merid. (1826-28) t. 53. 
Limo taurinus Rumph. Herb. Amb. 2: 105. 

The status of both Citrus bergamina Risso, as a species, and 
of Limo taurinus Rumph. is very doubtful. The reduction fol- 
lows Miquel and Hasskarl. I suspect that Limo taurinus is 
merely a form of Citrus hystrix DC. 

CITRUS MEDICA Linn. Sp. PL (1753) 782. 

Malum citrium Rumph. Herb. Amb. 2: 99, t. 25. 

This was placed by Loureiro, Fl. Cochinch. (1790) 465, under 
Citrus medica Linn., and by Hasskarl, Neue Schlussel (1866) 
31, under Citrus grandis Hassk. var. oblonga Hassk. 

CITRUS HYSTRIX DC. Cat. Hort. Monsp. (1813) 97. 

Limo tuberosus Rumph. Herb. Amb. 2: 101, t. 26, f. 1. 
Limo unguentarius Rumph. Herb. Amb. 2: 103. 
Limo agrestis Rumph. Herb. Amb. 2: 104, t. 27. 

The forms with the rugose fruits appear to be the same as 
the Philippine form described by Blanco as Citrus torosa. Limo 
tuberosus has been reduced to Citrus medica Linn, var., to C. 
hystrix DC. var., and to Citrus bergamia Risso var. unguen- 
taria Roem. Limo unguentarius has been reduced to Citrus 
medica Linn, var., to C. hystrix DC. var., and to C. bergamia 



298 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Risso var. unguentaria Roem. ; while Limo agrestis has been 
reduced to Citrus medica Linn, var., C. hystrix DC. var., C. 
bergamia Risso var. ventricosa Roem., Papeda rumphii Hassk., 
and Citrus papeda Miq. 

CITRUS AURANTIUM Linn. Sp. PL (1753) 782, var. 

Aurantium acidum Rumph. Herb. Amb. 2: 111, t. 33. 

This figure seems to represent one of the sour oranges. It 
has been reduced by various authors to Citrus fusca Lour., C. 
aurantium Linn. var. vulgaris Risso, C. vulgaris Risso, C. amara 
Hassk., and C. bigaradia Risso. 

To Citrus aurantium should probably also be referred 
Aurantium verrucosum Rumph., Herb. Amb. 2: 116, of Banda, and 
Aurantium pumilum madurense Rumph., 1. c, of Madura, and 
probably also the forms indicated as Aurantium acidum II and III 
on page 112. 

CITRUS NOBILIS Lour. Fl. Cochinch. (1790) 466. 
Aurantium sinense Rumph. Herb. Amb. 2: 113. 
Aurantium sinense II Rumph. Herb. Amb. 2: 113. 

This is the common loose-skinned orange commonly referred 
to Citrus nobilis Lour. Hasskarl, Neue Schliissel (1866) 33, 
refers A. sinense to Citrus nobilis var. melanocarpa Hassk. and 
A. sinense II to C. nobilis var. microcarpa Hassk. 

CITRUS sp. 

Limo ventricosus Rumph. Herb. Amb. 2: 102, t. 26, f. 2. 

The figure represents a form somewhat approaching the true 
lemon. It has been reduced by various authors to Citrus medica 
Risso, C. hystrix DC, C. aurantium Linn. var. limonum Risso, 
C. bergamia Risso var. ventricosa Roem., and C. limonum Risso. 

CITRUS sp. 

Limonellus aurarius Rumph. Herb. Amb. 2: 109, t. 30. 

- This has been reduced by various authors to Citrus aurantium 
Linn., C. limetta Risso, C. limetta var. auraria Risso, and C. 
hystrix DC. Its true position is very uncertain, but it may 
be a form of Citrus limetta Risso. 

CITRUS sp. 

Aurantium verrucosum Rumph. Herb. Amb. 2: 115, t. 35. 

This has been reduced by various authors to Citrus nobilis 
Lour., C. aurantium Linn., C. pompelmos var. racemosus Risso, 
C. decumana var. racemosa Roem., and C. decumana var. 
verrucosa Miq.; while Hasskarl, Neue Schliissel (1866) 33, sug- 



SIMARUBACEAE 



299 



gests that it may be Citrus macracantha Hassk. The figure 
might be either a form of Citrus aurantium Linn, or a small- 
fruited form of Citrus maxima (Burm.) Merr. (C. decumana 
Linn.). 

SIMARUBACEAE 

BRUCEA J. S. Miller 

BRUCEA AM ARISSI M A (Lour.) Merr. in Philip. Journ. Sci. 10 (1915) 
Bot. 18. 

Gonus amarissimus Lour. Fl. Cochinch. (1790) 658. 

Brncea sumatrana Roxb. Hort. Beng. (1814) 12 (type!), Fl. Ind. ed. 

2, 1 (1832) 449. 
Lussa radja Rumph. Herb. Amb. 7: 27, t. 15. 

This characteristic and widely distributed Malayan species 
is not represented in our Amboina collections. The Rumphian 
figure is the full basis of Brucea sumatrana Roxb. by citation in 
the original place of publication, Hort. Beng. (1814) 12. It is 
also cited by Loureiro in the original description of his Gonus 
amarissimus, Fl. Cochinch. (1790) 658. The form "II ex Solora" 
Rumph., 1. c. 28, may be referable to the same species; it was 
characterized as differing from the Javan form in its longer 
inflorescences and in its larger and more intensely bitter fruits. 

Al LANTH US * Desf ontaines 

AILANTHUS I NTEGRI FOLIA Lam. Encycl. 3 (1791) 417 (type!). 
Ailanthns moluccana DC. Prodr. 2 (1825) 89 (type!). 
Arbor coeli Rumph. Herb. Amb. 3: 205, t. 132. 

This species is not represented in our Amboina collections. 
Arbor coeli Rumph. is the whole basis of Ailanthus (Aylanthus) 
integrifolia Lam. as published in 1791, while A. moluccana DC, 
1825, is merely a new name for Lamarck's species, so that both 
must primarily be interpreted from the Rumphian figure and 
description. Linnaeus, Mant. 2 (1771) 379, thought that it 
might possibly be referable to Adenanthera falcataria Linn., 
which is Albizzia falcata (Linn.) Backer (see p. 249) ; while 
Henschel erroneously referred it to Connarus pentagijnus Lam. 

SA MADERA f Gaertner 
SAMADERA INDICA Gaertn. Fruct. 2 (1791) 352, t. 156. 
Lanius Rumph. Herb. Amb. 3: 194, t. 1M. 

This species is not represented in our Amboina collection. 

* Retained name, Vienna Code; Pongelion Adans. (1763) Is older, 
f Retained name, Vienna Code; Locandi Adans. (1763) is older. 



300 



RUMPHIUS'S HERBARIUM AMBOINENSE 



The figure is very poor and does not conform very well with 
Samadera indica Gaertn., although the plant described is cer- 
tainly referable to this genus. The only previously suggested 
reduction is Teysmann's opinion, quoted by Hasskarl, Neue 
Schlussel (1866) 66, that it was a species of Samadera. Botan- 
ical material from the Moluccas may show Lanius to be speci- 
fically distinct from Samadera indica Gaertn. 

SOULAMEA Lamarck 

SOULAMEA AMARA Lam. Encycl. 1 (1785) 449. 

Cardiocarpus amarus Reinw. Syll. Ratisb. 2 (1828) 14. 
Rex amaroris Rumph. Herb. Amb. 2: 129, t. hi. 

The Rumphian species was first reduced by Linnaeus to Ophio- 
xylon serpentinum Linn., in Stickman Herb. Amb. (1754) 9, 
Amoen. Acad. 4 (1759) 121, an entirely erroneous disposition 
of it. Lamarck cites Rex amaroris Rumph. as a synonym of 
Soulamea amara Lam. in the original description of the genus 
and species, the description being based on actual specimens 
from New Britain ; it is barely possible that the Moluccan form 
is specifically distinct from the New Britain one, but this can 
be determined only by a comparison of specimens from these 
localities. 

BURSERACEAE 

CAN AR I U M Linnaeus 

CANARIUM DECUMAN U M Gaertn. Fruct. 2 (1791) 99, t. 102. 

Pimela decumana Blume Mus. Bot. 1 (1850) 223. 
Canariopsis decumana Miq. Fl. Ind. Bat. 1 2 (1859) 652. 
Canarium decumanum Rumph. Herb. Amb. 2: 166, t. 55. 

This is not represented in our Amboina collections, and like 
several other species of the genus it must be interpreted largely 
from the description and the figure given by Rumphius. It is 
almost certain that the species as described by Engler, in DC. 
Monog. Phan. 4 (1883) 132, under Canarium decumanum is not 
the same as the plant that Rumphius described. The status 
of the species must await the results of further field work. 

CANARIUM LEGITI MUM (Blume) Miq. Fl. Ind. Bat. 1 2 (1859) 647. 

Pimela legitima Blume Mus. Bot. 1 (1850) 222. 

Dammara nigra legitima Rumph. Herb. Amb. 2: 162, t. 53. 

This is not represented in our Amboina collections. It is 
probable that Blume in describing the species correctly reduced 



BURSERACEAE 



301 



the Rumphian plant. Blume's species was based on actual 
specimens with a reference to the Rumphian name and figure. 

CANARIUM BALSAM I FERUM Willd. Sp. PL 4 2 (1804) 760 (type!). 

Boswellia balsamifera DC. Prodr. 2 (1825) 76 (type!). 
Pimela glabra Blume Mus. Bot. 2 (1850) 222 (type!). 
Canariopsis glabra Miq. FL Ind. Bat. 1 2 (1859) 653 (type!). 
Canarium odoriferum leve Rumph. Herb. Amb. 2: 156, t. 50. 

This is not represented in our Amboina collections. The 
status of the species is wholly doubtful, and all the names cited 
above must be interpreted entirely from Rumphius, as all are 
based solely on his description and figure. Like a number of 
other species of Canarium, the exact status of Canarium balsa- 
miferum Willd. must await further botanical exploration of 
Amboina; the only certain thing about the status of the species 
is that it is a true Canarium. 

CANARIUM COMMUNE Linn. Mant. 1 (1767) 127 (type!). 

Canarium mehenbethene Gaertn. Fruct. 2 (1791) 98, saltern quoad syn. 
Rumph. 

Canarium indicum Linn. Amoen. Acad. 4 (1759) 143 pro minore parte. 
Canarium moluccanum Blume Mus. Bot. 1 (1850) 216. 
Canarium vulgare Rumph. Herb. Amb. 2: 145, t. 47, (excl. /. E, 
F. G?) . 

Amboina, Binting and the town of Amboina, Robinson PI. Rumph. Amb. 
381, September 27, 1913, in flower; Gelala, Robinson PI. Rumph. Amb. 380, 
July 16, 1913, in fruit, locally known as kanari. 

Canarium vulgare Rumph. is the whole basis of Canarium 
commune Linn., as originally published by Linnaeus in his 
Mantissa 1 (1767) 127, and the species must be interpreted from 
the Rumphian reference. The species has been correctly inter- 
preted by practically all authors, as it is widely destributed in 
the Malay Archipelago and is a characteristic and well-known 
one. Canarium indicum Linn., not listed in Index Kewensis, is 
an older name, but I believe it should be abandoned for the reason 
that Linnaeus cited under it all the species of Canarium figured 
by Rumphius, t. U7 to 56, inclusive. He abandoned the name, 
apparently realizing later that numerous species were involved, 
and made the first plate, t. U7, the type of his Canarium com- 
mune. While the first plate cited under Canarium indicum, 
that is t. U7 , might be interpreted as the type of Canarium 
indicum Linn., it really represents that species only in small part. 
I can see no valid reason for considering Canarium moluccanum 
Blume other than a form of C. commune Linn. Figures E, F, 



302 



RUMPHIUS'S HERBARIUM AMBOINENSE 



and G, of plate U7 , probably represent merely variations of 
Canarium commune Linn. Hasskarl, Neue Schliissel (1866) 36, 
places them under C. commune Linn, and C. moluccanum Blume. 

CANARIUM HIRSUTUM Willd. Sp. PL 4 2 (1804) 760 (type!). 

Boswellia hirsuta DC. Prodr. 2 (1825) 76 (type!). 
Pimeia hirsuta Blume Mus. Bot. 1 (1850) 233 (type!). 
Canariopsis hirsuta Miq. Fl. Ind. Bat. 1 2 (1859) 653 (type!). 
Canarium odoriferum hirsutum Rumph. Herb. Amb. 2: 157, t. 51. 

This is not represented in our Amboina collections. Canarium 
hirsutum Willd. and all of the synonyms cited above must be 
interpreted wholly from Rumphius, on whose description and 
figure all are based. The probabilities are that Canarium his- 
pidum Blume will prove to be a synonym of it, but additional 
botanical material from Amboina will be necessary before the 
exact status of Canarium hirsutum Willd. can definitely be fixed. 

CANARIUM ACUTI FOLIUM (DC.) comb. nov. 

Marignia acutifolia DC. Prodr. 2 (1825) 79 (type!). 
Canarium nigrum Roxb. Fl. Ind. ed. 2, 3 (1832) 138. 
Pimeia acutifolia Blume Mus. Bot. 1 (1850) 221, excl. syn. Zipp. 
Dammara nigra Rumph. Herb. Amb. 2: 160, t. 52. 

Amboina, Mahija, Robinson PI. Rumph. Amb. 877, October 3, 1913, in 
light forests, altitude about 300 meters, locally known as nanari. 

Marignia acutifolia DC. was based entirely on the Rumphian 
reference cited above and must be interpreted from it and from 
Amboina material. After a careful study of the description as 
given by Rumphius, I am convinced that the specimen cited 
above, although with an abnormal fasciated inflorescence, rep- 
resents the plant described and in all probability the one figured, 
although the figure is crude and unsatisfactory. It is very 
certain that Canarium rostratum Zipp., referred here by Blume, 
is not the plant described or figured by Rumphius, but represents 
a distinct species apparently very closely allied to Canarium 
oleosum (Lam.) Merr. (C. microcarpum Willd.). It is to be 
noted that Blume, in citing Rumphius under Pimeia acutifolia, 
erroneously gives the reference as "Dammara nigra II. p. 160, 
t. 72," instead of Canarium nigrum Herb. Amb. 2: 160, t. 52, 
as it should be. Canarium nigrum Roxb., which is scarcely 
described by him, belongs here at least in part. It is not listed 
in Index Kewensis. He cites Dulcamara (sic!) nigra Rumph., 
Amb. II. 162. t. 52 and 53, as representing Canarium nigrum 
Roxb; but the two plates manifestly represent two distinct 
species, the latter being Canarium legitimum Blume (see p. 300) . 



BURSERACEAE 



303 



CANARIUM OLEOSUM (Lam.) Engl, in Engl. & Prantl Nat. Pflanzenfam. 
3 4 (1896) 241. 
Amyris oleosa Lam. Encycl. 1 (1783) 362 (type!). 
Canarium microcarpum Willd. Sp. PI. 4 2 (1805) 760 (type!). 
Nanarium minimum s. oleosum Rumph. Herb. Amb. 2: 162 (t. 54.?). 

Amboina, Mahija, Robinson PL Rumph. Amb. 376, October 3, 1913, in 
light forest at an altitude of 325 meters, locally known as nanari. 

Amyris oleosa Lam. and Canarium microcarpum Willd. are 
both based solely on Rumphius, and strictly must be interpreted 
by the Rumphian description and figure; they are, therefore, 
exact synonyms, and the older name is here retained. The 
reference of Nanarium oleosum Rumph. to Pimela oleosa Lour., 
Fl. Cochinch. (1790) 408, is a manifest error, as Loureiro's 
species was described from Cochin-China material that in all 
probability represents a species different from the Amboina one. 
Even though Loureiro cites the Rumphian name as a synonym 
and took his specific name from Rumphius, the reference to 
the Herbarium Amboinense should be excluded in interpreting 
his species. The present interpretation of Canarium oleosum 
(Lam.) Engl, follows the conventional interpretation of Cana- 
rium microcarpum Willd., and the Amboina specimen cited 
agrees closely with other botanical material from the Moluccas, 
so named, and with Rumphius's description ; it does not, however, 
agree well with the figure given by Rumphius, which may in- 
dicate some mixture between the Rumphian figure and descrip- 
tion. Canarium rostratum Zipp. should be critically compared 
with it. 

CANARIUM SYLVESTRE Gaertn. Fruct. 2 (1791) 99, t. 102. 

Canarium silvestre alterum Rumph. Herb. Amb. 2: 155, t. 49. 
Amboina, Hoetoemoeri road, Robinson PI. Rumph. Amb. 378, September 
30, 1913, in forests, altitude 400 meters; Hitoe messen, Robinson PI. Rumph. 
Amb. 379, October 14, 1913, in forests, altitude about 200 meters, locally 
known as dammara itam, nanari, nanari utan, and nanari puti daun alus. 

The specimens agree closely with both the description and 
figure given by Rumphius, and undoubtedly represent his Cana- 
rium silvestre alterum. While there may be some doubt as to 
whether or not the fruit figured by Gaertner represents the 
exact form described by Rumphius, still it seems best to retain 
Gaertner's name in its accepted application, especially in view 
of the fact that he cites the Rumphian figure and description 
as representing his species. The citation of Rumphius under 
Canarium sylvestre Gaertn. has been followed by all authors 
except Loureiro, who places it, with doubt, under Pimela nigra 



304 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Lour., Fl. Cochinch. (1790) 407, where it certainly does not 
belong. 

CANARIUM ZEPHYRINUM Blume Mus. Bot. 1 (1850) 217; Miq. FL 
Ind. Bat. 1 2 (1859) 643; March, in Baill. Adansonia 8 (1867-68) 
53; Engl, in DC. Monog. Phan. 4 (1883) 149. 
Canarium zephyrinum Rumph. Herb. Amb. 2: 151. t. U8. 

This species is not represented in our Amboina collections, 
unless it be a form of Canarium commune Linn., which seems 
to be very probable. The figure is rather crude, but certainly 
represents a form very closely allied to Canarium commune 
Linn., where it has been referred by many authors, including 
Willdenow, Poiret, Schultes, and Don. Canarium zephyrinum 
Blume is a species of doubtful status and in the latest monograph 
of the family is placed under Canarium among the "species 
dubiae." 

CANARIUM ZEYLANICUM (Retz.) Blume Mus. Bot. 1 (1850) 218. 

Amyris zeylanica Retz. Obs. 4 (1786) 25. 

Arbor zeylanica Rumph. Herb. Amb. 2: 153 (in Burm. obs.). 

The reduction follows Blume, Mus. Bot. 1 (1850) 218, as this 
is undoubtedly the correct disposition of Rumphius's Arbor 
zeylanica. 

CANARIUM PIMELA Konig in Konig & Sims Ann. Bot. 1 (1805) 361. 
Pimela nigra Lour. Fl. Cochinch. (1790) 407. 

Canarium nigrum Engl, in Engl. & Prantl Nat. Pflanzenfam. 3 4 (1896) 

240, non Roxb. 
Canarium sinense I Rumph. Herb. Amb. 2: 154? 

The identity of Canarium sinense I of Rumphius with Pimela 
nigra Lour. = Canarium pimela Konig is entirely problematical 
as the description is inadequate. I merely follow Blume, Mus. 
Bot. 1 (1850) 220, in this reduction. Rumphius's material was 
from China. 

CANARIUM ALBUM (Lour.) Rausch ex DC. Prodr. 2 (1825) 80. 
Pimela alba Lour. Fl. Cochinch. (1790) 408. 
Canarium sinense II Rumph Herb. Amb. 2: 154. 

The identity of Canarium sinense II with Pimela alba Lour, 
is more or less problematical, although Loureiro in the original 
publication of the species cites Rumphius, and the native names 
given by Rumphius agree with those cited by Loureiro. 
Rumphius's material was from China, not from Amboina. 
Engler, in DC. Monog. Phan. 4 (1883) 149, places Canarium 
album under the "species dubiae." * 

* See Guillaumin, in Lecomte Fl. Gen. Indo-Chine 1 (1911) 714, who 
gives a full description and a figure of this species. 



MELIACEAE 



305 



PI MELA CARYOPHYLLACEA Blume Mus. Bot. 2 (1850) 222. 
Canarium sinense III Rumph. Herb. Amb. 2: 154 (type!). 

Based originally on Chinese material by Rumphius. Pimela 
caryophyllacea Blume is based wholly on Rumphius's very brief 
description, and all that can be said regarding it is that it is 
probably a species of Canarium. 

CANARIOPSIS PAUCIJUGA Miq. Fl. Ind. Bat. 1 2 (1859) 653 (type!). 
Pimela paucijuga Blume Mus. Bot. 1 (1850) 226 (type!). 
Canarium odoriferum leve var. Rumph. Herb. Amb. 2: 156. 

A species of entirely doubtful status, based wholly on the 
Rumphian description and to be interpreted solely by it. All 
that can be definitely stated regarding it is that it is a species 
of Canarium. Its further determination must await additional 
botanical material from Amboina. 

GARUGA Roxburgh 

GARUGA ABILO (Blanco) Merr. in Govt. Lab. Publ. (Philip.) 35 (1905) 73. 
Guiacum abilo Blanco Fl. Filip. (1837) 364. 

Garuga mollis Turcz. in Bull. Soc. Nat. Mosc. 31 1 (1858) 475. 
Capraria Rumph. Herb. Amb. 2: 139? 

Capraria Rumph. is here tentatively referred to Garuga abilo 
Merr. The description applies closely, although Rumphius does 
not describe the flowers or the fruits. The only previously 
suggested reduction is Teysmann's opinion, quoted by Hasskarl, 
Neue Schlussel (1866) 35, that Capraria might belong in the 
Sapindaceae. 

PROTIUM Burman f. 

PROTIUM JAVANICUM Burm. f. Fl. Ind. (1768) 88. 
Amyris protium Linn. Mant. 1 (1767) 65. 
Tingulong Rumph. Herb. Amb. 7: 54, t. 23, f. 1. 

This reduction, manifestly the correct disposition of Tingu- 
long, was made by Burman f. in the original description of 
Protium javanicum Burm. f . 

MELIACEAE 

TOON A Roemer 

TOON A SURENI (Blume) comb. nov. 

Swietenia sureni Blume Cat. Gew. Buitenz. (.1823) 72. 
Cedrela febrifuga Blume Bijdr. (1825) 180. 
Surenus febrifuga O. Kuntze Rev. Gen. PL 1 (1891) 111. 
Surenus Rumph. Herb. Amb. 3: 66, t. 39. 

This is not represented in our Amboina collections. The 
illustration is unmistakably that of Cedrela, or Toona as the 

144971 20 



306 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Indo-Malayan representatives of the group are called by some 
botanists. More than one species may be included in the de- 
scription, Hasskarl, Neue Schliissel (1866) 51, definitely stating 
that three are represented; namely, Cedrela toona Roxb., C. 
febrifuga Blume, and C. inodora Hassk. Roxburgh was the 
first botanist to recognize the approximate position of Surenus 
and reduced it to Cedrela toona Roxb., in Fl. Ind. ed. 2, 1 (1832) 
635. The typical form of Roxburgh's species is confined to 
India, but a few varieties occur in the Malay Archipelago and 
in Australia.* Surenus alba Rumph., Herb. Amb. 3: 126, is con- 
sidered by Hasskarl, Neue Schliissel (1866) 51, to represent 
Cedrela inodora Hassk., which species C. de Candolle treats as a 
variety of Cedrela febrifuga Blume; Cedrela febrifuga Blume 
var. inodora (Hassk.) C. DC, Records Bot. Surv. Ind. 3 (1908) 
373. Hasskarl also refers to Cedrela febrifuga Blume the form 
described by Rumphius as Surenus rubra Rumph., Herb. Amb. 
3: 126, which is probably the correct disposition of it. The 
oldest specific name, that supplied by Swietenia sureni Blume, 
is here adopted. 

XYLOCARPUS Koenig 

XYLOCARPUS GRANATUM Koen. in Naturf. 20 (1784) 2. 

Carapa obovata Blume Bijdr. (1825) 179. 

Xylocarpus obovatus Juss. Mem. Mus. Paris 19 (1830) 244. 

Carapa indica Juss. in Diet. Sci. Nat. 7: 31. 

Carapa moluccensis Lam. Encycl. 1 (1785) 621 p. p., quoad Rumph. 

t. 61, excl. descr. 
Monosoma littorata Griff. Notul. 4 (1854) 502. 

Granatum litoreum III parvifolium Rumph. Herb. Amb. 3: 93, t. 61. 

This species is not represented in our Amboina collections. 
The description, but to a less degree the figure, given by 
Rumphius is clearly Xylocarpus {Carapa) obovatus Juss., as cur- 
rently interpreted, but Xylocarpus obovatus Juss. is apparently 
identical with the earlier Xylocarpus granatum Koen. In 
Rumphius's description the shape of the leaves, rounded at the 
apex, and the size of the fruits, indicated as larger than those of 
Granatum litoreum I latifolium, are the determining points. 

The synonymy between Xylocarpus granatum Koen. (X. obo- 
vatus Juss.) and X. moluccensis Lam. as interpreted below, is 
curiously confused. C. de Candolle f recognizes the two species 
Carapa obovata Blume and Carapa moluccensis Lam., but 
erroneously reduced Xylocarpus granatum Koen. to Carapa 

* See C. de Candolle, Records Bot. Survey India 3 (1908) 346. 
f Monog. Phan. 1 (1878) 718, 719. 



MELIACEAE 



307 



moluccensis Lam. Hiern * erred in referring both to a single 
species, Carapa moluccensis Lam. The type of Koenig's species 
was from the Tranquebar coast, India, and his description applies 
unmistakably to the form with obovate leaves and large fruits ; 
that is, the species that grows in the mangrove swamps; "folia 
* * * oblongo-clavata, rotundata;" "Habitat in silvis rhizo- 
phoreis." Lamarck, in the original description of Carapa 
'moluccensis, which was based wholly on "Granatum litoreum s. 
Martahul Rumph. Amb. 3. p. 92. Tab. 61," confused the two 
forms; the plate reference is to Xylocarpus granatum Koenig, 
but the description that he compiled was from Granatum lito- 
reum I latifolium Rumph. as indicated by the page reference 
and his description. He disposed of t. 62 of the Herbarium 
Amboinense, to which his description wholly applies, thus: 
"Variat foliis acutioribus. Ibidem t. 62." The two species 
are very strongly marked. Xylocarpus granatum Koen. grows 
in the mangrove swamps and has oblong to obovate leaflets, very 
large fruits, and a smooth, rather thin bark. Xylocarpus 
moluccensis Lam. grows on the open coasts and has usually ovate, 
acute leaflets ; much smaller fruits than X. granatum Koen. ; and 
a thick, very flaky bark. 

XYLOCARPUS MOLUCCENSIS (Lam.) M. Roem. Syn. Hesper. (1846) 
124 (type!). 

Carapa moluccensis Lam. Encycl. 1 (1785) 621, quoad descr., excl. 

Rumph. t. 61 (type!); C. DC. Monog. Phan. 1 (1878) 719, excl. 

syn. Koenig et Willdenow. 
Carapa rumphii Kostel. Allg. Med.-Pharm. Fl. 5 (1836) 1988 (type!). 
Xylocarpus carnulosus Zoll. & Mor. Nat. Geneesk. Arch. Neerl. Ind. 

2 (1845) 582, ex descr. 
Xylocarpus forstenii Miq. Ann. Mus. Bot. Lugd. Bat. 4 (1868) 62, ex 

descr. 

Granatum litoreum I latifolium Rumph. Herb. Amb. 3: 92, t. 62. 
Granatum litoreum III latissimum Rumph. Herb. Amb. 3: 92. 
Amboina, Amahoesoe, Robinson PI. Rumph. Amb. 491, September 19, 
1913, along the seashore. 

Lamarck, as noted above, in proposing the binomial Carapa 
moluccensis, based his description on Granatum litoreum I lati- 
folium, but erroneously referred to his species t. 61 of the Her- 
barium Amboinense, which is Xylocarpus granatum Koenig. 
His species, manifestly, should be interpreted by the form de- 
scribed, rather than by the figure that he erroneously referred 
to it. The rather confused synonymy between Xylocarpus gra- 
natum Koenig and X. moluccensis M. Roem., is here discussed 



* Hooker f. Fl. Brit. Ind. 1 (1875) 567. 



308 RUMPH IUS'S HERBARIUM AMBOINENSE 



under the former. The Rumphian illustration, t. 62, is the 
type and whole basis of Carapa rumphii Kostel. Xylocarpus 
carnulosus Zoll. & Mor., type from eastern Java, and X. forstenii 
Miq., type from Celebes, appear to be synonymous with Xylo- 
carpus moluccensis M. Roem. 

SAN DOR I CUM Cavanilles 

SANDORICUM KOETJAPE (Burm. f.) Merr. in Philip. Journ. Sci. 7 (1912) 
Bot. 237. 

Melia koetjape Burm. f. Fl. Ind. (1768) 101. 
Trichilia nervosa Vahl Symb. 1 (1790) 31. 
Sandoricum indicum Cav. Diss. 4 (1787) 359, t. 202, 203. 
Sandoricum domesticum Rumph. Herb. Amb. 1: 167, t. 64. 
Amboina, Robinson PL Rumph. Amb. 488, July 30, 1913, near the town 
of Amboina. 

The generic name Sandoricum was apparently taken from 
Rumphius, and the form figured and described by him has 
been consistently referred to Sandoricum indicum Cav., mani- 
festly a synonym of Melia koetjape Burm. f. Burman's specific 
name, being much the older, has been previously adopted by 
me, and under our rules of nomenclature it must be maintained. 
The forms briefly described by Rumphius as Sandoricum silvestre 
and Sandoricum Cajim Gulur are undoubtedly referable to Sando- 
ricum koetjape Merr., the former the spontaneous form with 
acid fruits, the latter with rather large, sweet fruits. Like most 
cultivated fruit trees, considerable variation is found in the 
characters of the fruit of the santol. 

DYSOXYLUM Blume 

DYSOXYLUM EUPHLEBIUM Merr. in Philip. Journ. Sci. 9 (1914) Bot. 
305. 

Alliaria Rumph. Herb. Amb. 2: 81, t. 20. 
Amboina, Hitoe messen, Robinson PL Rumph. Amb. 489, November 6, 
1913, in forests, altitude about 150 meters. 

This specimen certainly represents Alliaria Rumph., of which 
Rumphius figured a fruiting specimen. It is also apparently 
identical with the Philippine Dysoxylum euphlebium Merr., as 
a careful comparison between the specimen and the type of the 
latter species shows no essential differential characters. It is 
not Dysoxylum alliaceum Blume, which was described from 
Javan specimens. It was referred to Dysoxylum alliaceum 
Blume, Bijdr. (1825) 172, which disposition of it was accepted 
by Henschel, Walpers, Dietrich, and Miquel. Hamilton, Mem. 
Wern. Soc. 6 (1832) 305, placed it under Guarea alliaria Ham., 



MELIACEAE 



309 



the actual type of which, however, was a Bengal plant that 
is Dysoxylum hamiltonii Hiern. Hasskarl, Cat. Hort. Bot. 
Bogor. (1844) 221, erroneously reduced it to Hartighsea forsteri 
J uss.= Dysoxylum forsteri C. DC, a species of Australia and 
Polynesia; while Roemer, Hesper. (1844) 101, placed it under 
Prasoxylum alliaceum Roem., presumably a synonym of Dysoxy- 
lum alliaceum Blume, of Java and Sumatra. 

DYSOXYLUM sp. 

Arbor nussalavica Rumph. Herb. Amb. 7: 14, t. 8, f. 2. 

Manifestly this is a species of Dysoxylum, and one that should 
be readily recognized when once collected. Hamilton thought it 
was referable to the genus Guarea, and Teysmann, quoted by 
Hasskarl, Neue Schliissel (1866) 186, thought it was referable 
to Epicharis= Dysoxylum Blume. 

DYSOXYLUM sp. 

Arbor sebi Rumph. Herb. Amb. 7: 7. 

The plant described was from Java, there known as cadoja. 
This name is still applied to two or more species of Dysoxylum 
in Java, so that Arbor sebi is probably referable to this genus. 

LANSIUM (Rumphius) Correa 

LANSIUM DOMESTICUM Correa in Ann. Mus. Hist. Nat. Paris 10 (1807) 
157, t. 10. /. 1; Poir. in Lam. Encycl. Suppl. 3 (1813) 299; Jack in 
Trans. Linn. Soc. 14 (1823) 115, t. 4. 
Lansium Rumph. Herb. Amb. 1: 151, t. 5£. 

This common and well-known Malayan fruit tree is not rep- 
resented in our Amboina collections. Lansium was originally 
reduced, with doubt, to Averrhoa acida Linn, by Linnaeus, 
Amoen. Acad. 4 (1759) 119, in which he was followed by a few 
authors. Correa, Jack, Poiret, and recent authors generally, 
have referred it to Lansium domesticum, the correct disposition 
of it. The generic name Lansium is taken from Rumphius, 
and by some authors Rumphius is quoted as its author. The 
tree is commonly cultivated in most parts of the Malayan region, 
but like many other cultivated plants, it is poorly represented in 
herbaria. 

AGLAIA Loureiro 

AGLAIA ODORATA Lour. Fl. Cochinch. (1790) 173. 

Camunium sinense Rumph. Herb. Amb. 5: 26, t. 18, f. 1. 
Tsjiulang Rumph. Herb. Amb. 7: 38. 

This commonly cultivated ornamental tree is not represented 
in our Amboina collections, but the figure cited and both of the 



310 



RUMPHIUS'S HERBARIUM AMBOINENSE 



descriptions apply unmistakably to this well-marked species. 
Camunium sinense Rumph. was cited by Loureiro as a synonym 
of Aglaia odorata Lour, in the original description of that species, 
in which reduction he has been followed by practically all 
authors, as it is manifestly the correct disposition of it. Vitex 
pinnata Linn., Sp. PL (1753) 638, is cited in Index Kewensis 
as a synonym of Aglaia odorata Lour., but it is not clear on what 
grounds. It is based wholly on Pistacio-vitex Linn., Fl. Zeyl. 
No. Jfl5, the description of which is clearly a Vitex, certainly 
no meliaceous plant. Trimen, FL Ceyl. 3 (1895) 358, after 
examining Hermann's specimen, considers it to be a variety of 
Vitex altissima Linn. f. Druce, Bot. Exch. Club (1914) 413, 
has erroneously transferred Vitex pinnata Linn, to Aglaia as 
A. pinnata (Linn.) Druce, as the equivalent of Aglaia odorata 
Lour. Tsjiulang, as described by Rumphius in the Auctuarium 
is clearly the same as Camunium sinense Rumph., the common 
name of the latter being also tsjiulang. Hasskarl, Neue Schlus- 
sel (1866) 190, thought that it might be Aglaia odorata Lour, 
or Aglaia odoratissima Blume. 

AGLAIA SILVESTRIS (Roem.) comb. nov. § Euaglaia. 

Lansium silvestre Roem. Hesper. (1846) 99, ex Hassk. Neue Schlussel 

(1866) 20 (type!). 
Lansium silvestre Rumph. Herb. Amb. 1: 153, t. 55. 
Amboina, Hitoe lama, Robinson PL Rumph. Amb. U90 y October 11, 1913, 
in forests, altitude about 150 meters. 

Arbor circiter 16 m alta, ramulis et petiolis et inflorescentiis 
dense uniformiter cupreo- vel brunneo-lepidotis ; foliis circiter 
30 cm longis, foliolis circiter 12, suboppositis, oblongis, tenuiter 
acuminatis, vetustioribus utrinque glabris, usque ad 13 cm lon- 
gis; paniculis 20 ad 25 cm longis, anguste pyramidatis, multi- 
floris; floribus 5-meris, breviter pedicellatis, in ramulis ultimis 
racemose dispositis, calycis profunde obtuse 5-dentatis; tubo 
subellipsoideo, glabro, libero. 

A tree about 16 m high, the branchlets, petioles and rachis, 
and inflorescence densely and uniformly cupreous- or brownish- 
lepidote, the indumentum, however, not at all stellate-hairy. 
Leaves alternate, about 30 cm long, 6- or 7- jugate; leaflets 
firmly chartaceous to subcoriaceous, subopposite, oblong, 8 to 
13 cm long, 3 to 3.8 cm wide, rather pale when dry, shining, 
base acute to somewhat rounded, sometimes slightly inequi- 
lateral, apex slenderly long-acuminate, the acumen 1 to 1.5 
cm long, the lower surface browner than the upper, at maturity 
wholly glabrous; lateral nerves about 14 on each side of the 
midrib; petiolules 2 to 4 mm long. Panicles terminal and in 



MELIACEAE 



311 



the upper axils, narrowly pyramidal, 20 to 25 cm long, their 
peduncles 4 to 6 cm long, the lower branches spreading, 4 to 
6 cm long, the upper shorter. Flowers numerous, racemosely 
arranged on the ultimate branchlets, their pedicels 1 mm long 
or less, and with the calyces more or less lepidote. Calyx about 
1.4 mm long, prominently 5-toothed, the teeth rounded or obtuse, 
0.6 mm long. Petals 5, free, glabrous, oblong to elliptic-oblong, 
2.7 mm long, rounded. Staminal-tube free, glabrous, subellip- 
soid, contracted at the apex, the orifice small, round. Stamens 
5, the anthers attached near the base of the tube, about 1 mm 
long. 

Aglaia silvestris Merr., typified by the specimen cited above, 
is certainly the form described and figured by Rumphius as Lan- 
sium silvestre, the exact status of which has not been previously 
determined. Loureiro, Fl. Cochinch. (1790) 272, placed it under 
Quinaria lansium Lour. = Clausena punctata (Retz.) W. & A., 
a species that has little in common with the form Rumphius 
described. It was later referred by de Candolle, Prodr. 1 (1824) 
537, to Cookia punctata Retz., by reduction of Quinaria lansium 
Lour. Lansium silvestre Roem., Hesper. (1846) 99, is ap- 
parently merely a repetition of Rumphius's name. Teysmann, 
quoted by Hasskarl, Neue Schlussel (1866) 20, correctly placed 
it in the genus Aglaia. 

Its alliance is with Aglaia perviridis Hiern and Aglaia laxi- 
flora Miq., but it does not appear to be any of the numerous 
described forms and has been accordingly redescribed here as 
a new species. It is apparently the form described by Miquel as 
Aglaia ganggo Miq. forma amboinensis Miq., in Ann. Mus. Bot. 
Lugd.-Bat. 4 (1868) 47, but I consider it to be specifically distinct 
from Aglaia ganggo Miq., of which I have a series of excellent 
specimens. 

AGLAIA sp. 

Lansium montanum Rumph. Herb. Amb. 1: 154, t. 56. 

Probably an Aglaia, and one that should very readily be re- 
cognized when once collected, as Rumphius's figure presents a 
very characteristic species. Jack, Trans. Linn. Soc. 14 (1823) 
118, considered that it closely resembled his Milnea montana, of 
Sumatra = A #Zam; and Roemer, Hesper. (1846) 99, placed it 
under Selbya montana Roem., perhaps based on Rumphius's Lan- 
sium montanum, perhaps based on Milnea montana Jack (ori- 
ginal publication not seen by me) . Teysmann, quoted by Hass- 
karl, Neue Schlussel (1866) 20, considers it to be a species of 
Aglaia, which is probably the correct disposition of it, although 



312 



RUMPHIUS'S HERBARIUM AMBOINENSE 



it might possibly be a small-flowered species of Dysoxylum. In 
the plate the fruits are drawn entirely out of proportion to the 
leaves according to the measurements given by Rumphius. 

MELIACEAE OF UNCERTAIN STATUS 

Vidoricum silvestre Rumph. Herb. Amb. 1: 173, t. 67. 

This disposition of Vidoricum silvestre is suggested chiefly because the 
seeds, as figured by Rumphius, almost certainly pertain to some meliaceous 
plant. The description of the species is indefinite, and from it alone no 
rational idea of the plant can be obtained. 

DICHAPETALACEAE 

DICHAPETALUM Thouars 

DICH APETALU M MOLUCCANUM sp. nov. 

Funis butonicus minor Rumph. Herb. Amb. 5: 77, t. 4.1, f. 2. 
Amboina, Wae, Robinson PI. Rumph. Amb. 602 (type), November 26, 
1913, in light forests, altitude about 20 meters. 

Frutex scandens, ut videtur dioicus, ramulis junioribus inflor- 
escentiisque leviter cinereo-pubescentibus, foliis subtus ad costa 
nervisque plus minusve ciliato-hirsutis ; foliis oblongis, charta- 
ceis, usque ad 16 cm longis, integris, utrinque subaequaliter an- 
gustatis, basi acutis, apice breviter acuminatis, nervis utrinque 
circiter 7, subtus prominentibus, in siccitate nigrescentibus ; 
cymis axillaribus, solitariis, pedunculatis, usque ad 2 cm longis, 
laxis, paucifloris; floribus $ 5-meris, circiter 3 mm diametro, 
sepalis extus parce pubescentibus, petalis oblongis, truncatis, 
vix retusis, glabris, glandulis pubescentibus. 

A scandent shrub. Branches terete, glabrous, slender, 
brownish or somewhat reddish-brown, lenticellate, the branchlets 
similar in color, slightly pubescent with pale, scattered, usually 
appressed hairs. Leaves chartaceous, oblong, 11 to 16 cm long, 
4 to 6 cm wide, entire, subequally narrowed to the acute, equi- 
lateral, or rarely slightly inequilateral base, and to the distinctly 
but rather shortly blunt-acuminate apex, brown or olivaceous- 
brown when dry, the upper surface shining, quite glabrous, 
the lower surface of nearly the same color, ciliate-hirsute with 
scattered, spreading, pale hairs on the midrib, and to a less 
extent on the primary nerves, reticulations, and sometimes even 
the margins ; lateral nerves about 7 on each side of the midrib, 
prominent, curved, blackish when dry, faintly anastomosing, 
the reticulations distinct; petioles about 4 mm long, sparingly 
grayish-pubescent, ultimately glabrous. Cymes axillary, soli- 
tary, slightly grayish-pubescent with short appressed hairs, 



EUPHORBIACEAE 



313 



peduncled, about 2 cm long and wide, dichotomous, rather lax, 
few-flowered, the peduncle about as long as or slightly exceeding 
the petiole. Male flowers pale yellow, about 3 mm in diameter, 
their pedicels pubescent, 1 to 2 mm long, the bracteoles very 
small. Calyx about 2 mm long, externally cinereous-pubescent 
with short, appressed hairs, the lobes 5, oblong-ovate, obtuse, 
about 1.5 mm long. Petals glabrous, oblong, apex rounded- 
subtruncate, entire or very obscurely notched, not retuse or split, 
nearly 1.5 mm long. Stamens 5, about 1.5 mm long, glabrous. 
Glands densely pubescent, small. Pistillate or perfect flowers 
not seen. 

This species is apparently closely allied to Dichapetalum timo- 
riense (DC.) Engl., from which it is distinguished by certain 
floral characters, the flowers dioecious (or polygamous?), the 
glands densely pubescent, not glabrous. It seems to be even 
more closely allied to Dichapetalum papuanum (Becc.) Engl., 
of New Guinea, but has rather smaller, differently shaped leaves, 
while the petals are truncate and entire, rarely minutely notched, 
not at all bifid at the apex. 

Rumphius's description and figure agree sufficiently well with 
the specimen cited above to warrant the reduction of Funis 
butonicus minor to Dichapetalum moluccanum Merr. The only 
previously suggested reductions are Burman's opinion that it 
represented some species in the Contortae and Lamarck's opinion 
that it was near Menispermum. 

EUPHORBIACEAE 

PHYLLANTHUS Linnaeus 
PHYLLANTHUS NIRURI Linn. Sp. PL (1753) 981. 

Herba moeroris I alba Rumph. Herb. Amb. 6: 41, t. 17, f. 1. 
Amboina, Batoe merah, Robinson PL Rumph. Amb. 350, July 20, 1913, 
in ditches at low altitudes. 

This reduction was made by Linnaeus, in Stickman Herb. 
Amb. (1754) 26, Amoen. Acad. 4 (1759) 134, Syst. ed. 10 (1759) 
1264, Sp. PI. ed. 2 (1763) 1392, has been accepted by all authors 
who have had occasion to cite the Rumphian figure, and is cer- 
tainly the correct disposition of Herba moeroris I alba. 

PHYLLANTHUS URINARIA Linn. Sp. PI. (1753) 982. 

Herba moeroris II rubra Rumph. Herb. Amb. 6: 42, t. 17, /. 2. 
Amboina, Hatiwe ketsjil, Robinson PI. Rumph. Amb. 351, July 20, 1913, 
on coral limestone at low altitudes. 

This reduction was made by Linnaeus in the same publica- 
tions as those in which Herba moeroris I alba was reduced to 



314 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Phyllanthus niruri Linn., has been accepted by all authors, and 
is certainly the correct disposition of it. 

PHYLLANTHUS EMBLICA Linn. Sp. PL (1753) 982. 

Mirobalanus embilica Rumph. Herb. Amb. 7: 1, t. 1. 

This species is not represented in our Amboina collections. 
The reduction was first made by Linnaeus, in Amoen. Acad. 4 
(1759) 136, Syst. ed. 10 (1759) 1265, Sp. PL ed. 2 (1763) 1393, 
which is certainly the correct disposition of it. Rumphius notes 
that it was an introduced plant in Amboina. 

CI CCA Linnaeus 
CI CCA ACIDA (Linn.) comb. nov. 

Averrhoa acida Linn. Sp. PL (1753) 428. 
Cicca disticha Linn. Mant. 1 (1767) 124. 
Cicca nodiflora Lam. Encycl. 2 (1786) 1. 

Phyllanthus distichus Muell.-Arg. in DC. Prodr. 15 2 (1866) 413. 
Phyllanthus cicca Muell.-Arg. in Linnaea 32 (1863) 90. 
Cicca acidissima Blanco Fl. Filip. (1837) 700. 
Phyllanthus acidissimus Muell.-Arg. in Linnaea 32 (1865) 50. 
Phyllayithus cheramela Roxb. Hort. Beng. (1814) 104. 
Phyllanthus acidus Skeels in U. S. Dept. Agr. Bur. PL Ind. Bull. 148 
(1909) 17. 

Cheramela Rumph. Herb. Amb. 7: 34, t. 17, /. 2. 

This widely cultivated tree is not represented in our Amboina 
collections. The reduction of Cheramela Rumph. to Averrhoa 
acida Linn, was first made by Linnaeus, Amoen. Acad. 4 (1759) 
136, Sp. PI. ed. 2 (1762) 613, in which he was followed by Bur- 
man f. and Lamarck. Willdenow, followed by numerous other 
authors, referred it to Cicca disticha Linn. Miquel, Fl. Ind. 
Bat. 1 2 (1859) 372, referred it to Cicca nodiflora Lam., while 
Henschel and Pritzel referred it to Phyllanthus cheramela Roxb. 
All of these names are synonyms of Cicca acida (Linn.) Merr., 
Averrhoa acida Linn, being a synonym of Cicca disticha Linn. 
{Phyllanthus distichus Muell.-Arg.), but much earlier. I prefer 
to follow Robinson * in retaining Cicca as a genus distinct from 
Phyllanthus. 

BREYNIA Forster 

BREYNIA CERNUA (Poir.) Muell.-Arg. in DC. Prodr. 15' (1866) 439. 

Phyllanthus cernuus Poir. in Lam. Encycl. 5 (1804) 298. 

Melanthesa cernua Decne. in Nuov. Ann. Mus. Paris 3 (1834) 483. 

Aalius parvifolia Rumph. Herb. Amb. 3: 207. 
Amboina, Batoe gadjah and Negri lama, Robinson PL Rumph. Amb. U99, 
339, August 5 and October 8, 1913, in light forests, altitude 175 to 250 
meters, locally known as kartu utan. 

* Philip. Journ. Sci. 4 (1909) Bot. 87. 



EUPHORBIACEAE 



315 



The description is that of a Breynia in all respects, and 
undoubtedly Aalius parvifolia Rumph. is the same as Breynia 
cernua Muell.-Arg. The red, accrescent calyx is very character- 
istic. The only other suggested reduction of Rumphius's species 
is Hasskarl's, Neue Schliissel (1866) 67, reference of it to 
Sauropus albicans Rlume, which is certainly incorrect. 

Aalius latifolia Rumph., Herb. Amb. 3 : 207, very briefly de- 
scribed, is compared with Aalius parvifolia, having leaves two 
to three times as long as the latter. Hasskarl, Neue Schliissel 
(1866) 67, surmises that it may be a Glochidion. Its status is 
quite indeterminable from the data given by Rumphius. 

BACCAUREA Loureiro 
BACCAUREA NAN I H U A sp. nov. 

Nani hua Rumph. Herb. Amb. 3: 21, t. 9. 
Amboina Koesoekoesoe sereh and Mahija, Robinson PI. Rumph. Amb. 
831 (type), 330, October 3 and August 7, 1913, in light forest, altitude 
about 250 meters, locally known as haharu and as makarlasi. 

Arbor circiter 15 m alta, inflorescentiis dense ferrugineo- 
pubescentibus ; foliis oblongo-ovatis, coriaceis, glabris, in sicci- 
tate brunneis, usque ad 16 cm longis, integris, basi rotundatis, 
apice obtusis ad latissime obtuseque acuminatis, nervis utrinque 
circiter 9, subtus valde prominentibus ; racemis solitariis vel 
binis, 3 ad 5 cm longis, e ramis infra foliis, paucifloris, omnibus 
partibus dense ferrugineo-pubescentibus, sepalis oblongis, cir- 
citer 3 mm longis; fructibus depresso-globosis, 2 ad 2.3 cm 
diametro, extus ferrugineo-pubescentibus vetustioribus glabres- 
centibus, brunneis, pericarpio crassissimo. 

A tree about 15 m high, the very young branchlets and petioles 
slightly pubescent, the inflorescence densely ferruginous-pubes- 
cent, otherwise glabrous. Branches and branchlets brown or 
reddish-brown, terete. Leaves coriaceous, brown and somewhat 
shining when dry, oblong-ovate, entire, 11 to 16 cm long, 5 to 8 
cm wide, base rounded, narrowed upward to the obtuse or very 
broadly blunt-acuminate apex, the lower surface paler than the 
upper; lateral nerves about 9 on each side of the midrib, very 
prominent on the lower surface, curved-anastomosing, the re- 
ticulations distinct; petioles 2.5 to 3.5 cm long. Pistillate ra- 
cemes solitary or in pairs, from the branches below the leaves, 
few-flowered, 3 to 5 cm long, simple, all parts densely ferrugi- 
nous-pubescent. Pedicels about 4 mm long, jointed at about 
the middle, here supplied with a single, broadly ovate, 1.5 mm 
long bracteole. Sepals 5, oblong to oblong-ovate, subequal, sub- 
acute or obtuse, about 3 mm long, densely pubescent on both 



316 



RUMPHIUS'S HERBARIUM AMBOINENSE 



surfaces. Ovary ovoid, densely pubescent. Fruits brown when 
dry, depressed-globose, 2 to 2.5 cm in diameter, normally 3-celled, 
in cross-section with three very broadly rounded angles, more 
or less ferruginous-pubescent, in age becoming nearly glabrous, 
the pericarp very thick, inside reddish-brown when dry, some- 
what spongy in texture. 

Nani hua Rumph. was reduced by Loureiro to Baccaurea 
rami flora Lour., Fl. Cochinch. (1790) 661, the type of the genus, 
but Baccaurea ramiflora Lour, was actually described from speci- 
mens taken from cultivated trees in Cochin-China and is not 
the same as the Amboina Nani hua of Rumphius. Henschel, 
Poiret, and Pritzel accepted Loureiro's reduction of Nani hua. 
DeVriese and Poiret, in Lam. Encycl. 4 (1798) 419, thought 
that it might be a species of Eugenia, while Hasskarl, Neue 
Schlussel (1866) 47, quotes Teysmann's opinion that it might 
belong in the Myrtaceae. 

Baccaurea nanihua Merr. closely resembles Baccaurea philip- 
pinensis Merr. and Baccaurea hracteata Muell.-Arg. and mani- 
festly belongs in the same group as these two species. It is 
readily distinguished from both, however, by its more numer- 
ously nerved leaves. 

ANTIDESMA Burman 

ANTIDESMA BUNIUS (Linn.) Spreng. Syst. 1 (1825) 826. 
Stilago bunius Linn. Mant. 1 (1767) 122. 

Antidesma rumphii Tul. in Ann. Sci. Nat. Ill 15 (1851) 238 (type). 

Bunius sativa s. domestica Rumph. Herb. Amb. 3: 204, t. 131. 

Bunius agrestis Rumph. Herb. Amb. 3: 205, t. 131, f. A. 
Amboina, Robinson PL Rumph. Amb. 33h, September 13, 1913, from 
cultivated trees in the town of Amboina, locally known as kuti kata and 
kata kuti. 

Bunius sativus Rumph. (B. domestica Rumph.) was reduced 
by Linnaeus to Stilago bunius Linn, in the original description 
of that species, which, as Antidesma bunius (Linn.) Spreng., is 
certainly the correct disposition of it. Antidesma rumphii Tul. 
was based wholly on Bunius agrestis Rumph., which seems to 
me to be merely the spontaneous or subspontaneous form of 
Antidesma bunius Spreng.; I have accordingly here reduced 
Antidesma rumphii Tul. to Antidesma bunius (Linn.) Spreng. 

ANTIDESMA STIPULARE Blume Bijdr. (1826) 1125. 

Antidesma amboinense Miq. Ann. Mus. Bot. Lugd. Bat. 1 (1864) 218. 

Arbor nuda Rumph. Herb. Amb. 3: 89, t. 59. 
Amboina, Lateri, Batoe merah River, and vicinity of the town of Amboina, 
Robinson PI. Rumph. Amb. 355, 356, August, September, and October, 1913, 
in light woods, altitude 40 to 150 meters, locally known as kata kuti kam- 



EUPHORBIACEAE 



317 



bing ; probably referable here also is Rel. Robins. 1709, from Hitoe messen, 
Amboina, November 6, 1913, in forests, altitude about 150 meters, with 
smaller, narrower stipules and apparently more fleshy, slightly larger 
fruits than the other specimens. 

The Amboina specimens, typical Antidesma amboinense Miq., 
differ from material of Antidesma stipulare Blume from Nusa 
Kambangan, the type locality of Blume's species, and from Java 
in some details, notably in their larger, differently shaped sti- 
pules, but I have followed J. Mueller in the reduction of Anti- 
desma amboinense Miq. to Antidesma stipulare Blume. Arbor 
nuda Rumph. was reduced to the genus Antidesma by Teysmann, 
as quoted by Hasskarl, Neue Schlussel (1866) 54. 

The details of the figure are not good, the leaves being rep- 
resented as with but 5 to 8 pairs of nerves, while in the Amboina 
specimens there are usually about 15 pairs of nerves, and the 
characteristic stipules are not shown at all. However, these are 
indicated in the description thus: ' 'folia * * * suprema 
vero prope suum ortum unum alterumve gerunt foliorum seu 
squamulas." In spite of the discrepancies between the figure 
and the specimens cited above, I am confident that Arbor nuda 
Rumph. is here correctly interpreted, although some future mo- 
nographer may prefer to reinstate Miquel's Antidesma amboin- 
ense as a species distinct from the Javan Antidesma stipulare 
Blume. 

CROTON Linnaeus 

CROTON TIGLIUM Linn. Sp. PI. (1753) 1004. 

Tiglium officinale Klotz. in Nov. Act. Acad. Nat. Cur. 19 (1843) 

Suppl. 1: 418. 
Granum moluccanum Rumph. Herb. Amb. 4: 98, t. U2. 

This is not represented in our Amboina collections. The 
reduction of Granum moluccanum to Croton tiglium was first 
made by Linnaeus, in Stickman Herb. Amb. (1754) 16, Amoen. 
Acad. 4 (1759) 126, Syst. ed. 10 (1759) 1275, Sp. PL ed. 2 (1763) 
1426, which is the correct disposition of it, and which has been 
accepted by all authors. 

MICROCOCCA Bentham 

MICROCOCCA M ERCURI ALIS (Linn.) Benth. in Hook. Niger Fl. (1849) 
503. 

Tragia mercurialis Linn. Sp. PI. (1753) 980. 

Urtica mortua Rumph. Herb. Amb. 6: 49, t. 20, /. 2? 

Nothing resembling this species is represented in our Amboina 
collections. The reduction was first made by Linnaeus, in Stick- 
man Herb. Amb. (1754) 26, Amoen. Acad. 4 (1759) 134, Syst. 



318 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ed. 10 (1759) 1264, which, as Micrococca mercurialis Benth., 
is possibly the correct disposition of it. The species, however, is 
not known from the Moluccas. Pax and K. Hoffmann, in Engl. 
Pflanzenreich 63 (1914) 133, give its range as tropical Asia and 
Africa, extending to the southeast only as far as Malacca. 

M ALLOT US Loureiro 
MALLOTUS TILLIFOL1US (Blume) Muell.-Arg. in Linnaea 34 (1865) 190. 
Rottlera tiliifolia Blume Bijdr. (1825) 607. 
Halecus litorea Rumph. Herb. Amb. 3: 196, t. 126. 
Amboina, Robinson PL Rumph. Amb. 367, 368, August 8, 1913, along 
the seashore near the town of Amboina, locally known as baru laut. 

Halecus litorea Rumph. was originally reduced by Linnaeus 
to Croton aromaticus Linn., in Stickman Herb. Amb. (1754) 14, 
Amoen. Acad. 4 (1759) 125, in which entirely erroneous reduc- 
tion he was followed by Loureiro, Murray, Willdenow, Persoon, 
Henschel, and Miquel. Lamarck, Encycl. 2 (1786) 206, cites it 
under Croton tiliifolius Lam. var. aromaticus Lam., the type of 
the species being a specimen from Mauritius, the variety being 
merely a reduction of Croton aromaticus Linn. Teysmann's 
reduction of it to Schmidelia, as quoted by Hasskarl, Neue 
Schlussel (1866) 66, is entirely wrong and was probably due 
to some error in transcribing Teysmann's notes. 

MELANOLEPIS Reichenbach f. and Zollinger 

MELANOLEPIS M ULTIGLANDULOSA (Reinw.) Reichb. f. & Zoll. in 
Linnaea 28 (1856) 324. 
Croton multiglandulosus Reinw. ex Blume Cat. Gew. Buitenz. (1823) 
105. 

Rottlera multiglandulosa Blume Bijdr. (1825) 609. 
Melanolepis calcosa Miq. Fl. Ind. Bat. 1 2 (1859) 399. 
Mallotus moluccanus Muell.-Arg. in Linnaea 34 (1863) 185, non Croton 
moluccanus Linn. 

Melanolepis moluccana Pax & K. Hoffm. in Engl. Pflanzenreich 63 

(1914) 142, non Croton moluccanus Linn. 
Folium calcosum Rumph. Herb. Amb. 4: 129, t. 6A. 
Folium calcosum II Rumph. Herb. Amb. 4: 130. 
Amboina, Hitoe messen and Ayer putri, Robinson PI. Rumph. Amb. 359, 
360, August and October, 1913, on forested limestone hills and along road- 
sides, altitude 5 to 150 meters, locally known as kayu kapor. 

Folium calcosum Rumph. was reduced by Roxburgh, Fl. Ind. 
ed. 2, 3 (1832) 690, to Ricinus dicoccus Roxb., the description 
being based on specimens originating in Amboina and cultivated 
in the botanic garden at Calcutta ; it is a synonym of Melanolepis 
multiglandulosa Reichb. f. & Zoll. Miquel, Fl. Ind. Bat. I 2 
(1859) 399, reduced Ricinus dicoccus Roxb. together with 



EUPHORBIACEAE 



319 



Folium calcosum Rumph. to Melanolepis calcosa Miq., which is 
also a synonym of Melanolepis multiglandulosa Reichb. f. & Zoll. 
Hasskarl, Neue Schlussel (1866) 85, reduced Folium calcosum II 
to Melanolepis multiglandulosa Reichb. f. & Zoll. 

In regard to the proper specific name for this widely distri- 
buted and well-known species, I cannot follow J. Mueller in 
calling it Mallotus moluccanus Muell.-Arg. or Pax and K. Hoff- 
mann in calling it Melanolepis moluccana Pax & K. Hoffm., for 
the reason that nothing in the Linnean description of Croton 
moluccanus Linn., the name-bringing synonym, applies to this 
species. It was based on two references ; the first, Fl. Zeyl. SJf.6, 
and the second, Nux juglans moluccana bifida Burm. Fl. Zeyl. 
170. The first reference, Fl. Zeyl. 36^, is Givotia rottleriformis 
Griff., according to the actual specimens, leaves only, in Her- 
mann's herbarium.* The second reference, from which Lin- 
naeus took his specific name, is Aleurites moluccana (Linn.) 
Willd. without the slightest doubt. The specimen in Linnaeus's 
herbarium, quoted by J. Mueller, in DC. Prodr. 15 2 (1866) 958, 
under Mallotus moluccanus Muell.-Arg., is manifestly not the 
type and should be ignored in interpreting the Linnean species. 

MACARANGA Thouars 

MACARANGA MAPPA (Linn.) Muell.-Arg. in DC. Prodr. 15 2 (1866) 1000. 
Ricinus mappa Linn, in Stickman Herb. Amb. (1754) 14, Amoen. Acad. 
4 (1759) 124, Syst. ed. 10 (1759) 1277, Sp. PI. ed. 2 (1763) 1430 
(type!). 

Acalypha mappa Willd. Sp. PI. 4 (1805) 526 (type!). 
Mappa moluccana Spreng. Syst. 3 (1826) 878 (type!). 
Tanarius mappa O. Ktze. Rev. Gen. PL 2 (1891) 620 (type!). 
Folium mappae Rumph. Herb. Amb. 3: 172, t. 108. 

Amboina, Wae, Robinson PL Rumph. Amb. 361, November 26, 1913, in 
light forest, altitude about 20 meters, locally known as hahuhun. 

Folium mappae Rumph. is the whole basis of Ricinus mappa 
Linn., and thus typifies all the synonyms cited above, for Mappa 
moluccana Spreng. is merely a new name for Ricinus mappa 
Linn. Burman f., Lamarck, Poiret, Roxburgh, and Pritzel fol- 
lowed Linnaeus in considering it as Ricinus mappa Linn., but 
other authors have followed Willdenow and Sprengel and have 
placed it under Acalypha and Mappa. 

Macaranga mappa (Linn.) Muell.-Arg. is an imperfectly 
known species. In de Candolle's Prodromus 15 2 (1866) 1000 
J. Mueller apparently compiled his description largely, if not 
wholly, from Rumphius and erroneously reduced Croton grand- 

* See Trimen Fl. Ceyl. 4 (1898) 50. 



320 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ifolius Blanco to Macaranga mappa Muell.-Arg. as a synonym. 
Pax and K. Hoffmann, in Engl. Pflanzenreich 63 (1914) 320, 
assumed that Mueller was correct in his reduction of Blanco's 
species, but drew up their description and prepared a figure 
of Macaranga mappa wholly from Philippine material. The 
Philippine material cited by these authors has nothing whatever 
to do with the Amboina Macaranga mappa Muell.-Arg., which 
is so entirely distinct from Macaranga grandifolia (Blanco) 
Merr. that it must be placed in a different section of the genus. 
In vegetative characters, in its staminate inflorescence, and espe- 
cially in its pistillate inflorescence, Macaranga grandifolia 
(Blanco) Merr. is totally different from Macaranga mappa 
Muell.-Arg. The whole description and the figure given by 
Pax and Hoffmann go with Macaranga grandifolia (Blanco) 
Merr., together with the synonyms Macaranga porteana Andre, 
Mappa porteana Wats., and Croton grandifolius Blanco. 

MACARANGA HISPIDA (Blume) Muell.-Arg'. in DC. Prodr. 15 2 (1866) 
990. 

Mappa hispida Blume Bijdr. (1825) 624. 

Halecus rugosa Rumph. Herb. Amb. 3: 198. 
Amboina, Mahija, Robinson PL Rumph. Amb. 363, 36k, August 12, 1913, 
in light forest at an altitude of 250 meters, locally known as haleki and as 
bilang kinar. 

No previous reduction of Halecus rugosa Rumph. has been 
suggested. Rumphius's description agrees perfectly with the 
specimens cited above, which are at the same time apparently 
typical Macaranga hispida Muell.-Arg., which was originally 
described by Blume from Moluccan specimens. The Philippine 
material referred here is very much more pubescent than the 
Amboina specimens. 

MACARANGA TANARIUS (Linn.) Muell.-Arg. in DC. Prodr. 15 2 (1866) 
997. 

Ricinus tanarius Linn, in Stickman Herb. Amb. (1754) 14, Amoen. 
Acad. 4 (1759) 125, Syst. ed. 10 (1759) 1277, Sp. PI. ed. 2 (1763) 
1430 (type!). 
Mappa tanarius Blume Bijdr. (1825) 624. 
Ricinus mappa Roxb. Fl. Ind. ed. 2, 3 (1832) 690, non Linn. 
Tanarius minor alba Rumph. Herb. Amb. 3: 190, t. 121. 
Tanarius minor rubra Rumph. Herb. Amb. 3: 190. 
Amboina, Hoenoet, Robinson PL Rumph. Amb. 362, October 18, 1913, in 
wooded glens, altitude about 200 meters, locally known as hunua. 

The Rumphian figure and description are the whole basis of 
Ricinus tanarius Linn, and are hence the type of the species. 
Most early authors considered it under the Linnean binomial, 



EUPHORBIACEAE 



321 



Ricinus tanarius Linn., which in modern literature appears as 
Macaranga tanarius Muell.-Arg. Spanoghe referred Tanarius 
minor Rumph. to Mappa glabra Juss., which Pax and Hoffmann 
retain as a distinct species, Macaranga glabra (Juss.) Pax & 
Holfm. It is to be noted that the Amboina specimen, cited 
above, which agrees closely with the Rumphian description and 
with the greatly reduced figure, is by no means identical with 
Macaranga tanarius Muell.-Arg. as currently interpreted. The 
leaves are larger, glabrous, the petioles longer, and the stami- 
nate inflorescences are lax, long peduncled, up to 30 cm wide, 
and including the peduncle about 40 cm long. It is typical 
Macaranga tanarius (Linn.) Muell.-Arg., and it seems very 
probable that a critical revision of the group must lead to the 
adoption of another specific name for the widely distributed 
form that appears in herbaria under the name Macaranga 
tanarius Muell.-Arg. Hasskarl, Neue Schlussel (1886) 65, sug- 
gested that Tanarius minor rubra Rumph. might be the same as 
Mappa denticulata B\ume=Macaranga denticulata Muell.-Arg., 
but the distribution of the latter precludes the correctness of 
this reduction. I consider it to be a color variant of Macaranga 
tanarius Muell.-Arg. 

MACARANGA INVOLUCRATA (Roxb.) Baill. Etud. Gen. Euphorb. (1858) 
432. 

Urtica involucrata Roxb. Hort. Beng. (1814) 47, nomen nudum, Fl. 

Ind. ed. 2, 3 (1832) 592. 
Halecus terrestris vulgaris Rumph. Herb. Amb. 3: 197, t. 127. 
Halecus terrestris alba Rumph. Herb. Amb. 3: 198, t. 127 bis. 
Amboina, Batoe merah, Robinson PL Rumph. Amb. S65, 366, August 11, 
1913, hillsides at an altitude of about 20 meters, the former with pistillate 
flowers and immature fruits, the latter with staminate flowers. 

Halecus terrestris Rumph. was originally reduced by Linnaeus, 
with doubt, to Croton lacciferum Linn., in Stickman Herb. Amb. 
(1754) 14, Amoen. Acad. 4 (1759) 125, and following this reduc- 
tion, was placed by Willdenow, Sp. PL 4 (1805) 590, under Aleu- 
rites laccifera Willd. Miquel, Fl. Ind. Bat. I 2 (1859) 406, placed 
it, with doubt, under Acalypha bracteata Miq. The specimens 
cited above agree closely with the description and figure given 
by Rumphius and are, moreover, typical Macaranga involucrata 
Baill., a species known only from Amboina except for the plants 
cultivated in the botanic gardens at Calcutta and Buitenzorg. 

As representing Halecus terrestris alba Rumph., I cite PI. 
Rumph. Amb. 357 , 358, from Amahoesoe and the town of Am- 
boina, August, 1913, both with staminate flowers. The leaves 
are somewhat more pubescent and are less toothed than in the 

144971 21 



322 



RUMPHIUS'S HERBARIUM AMBOINENSE 



two specimens cited above, yet they are apparently referable to 
the same species. The Rumphian figure is poor, but the short 
description applies closely to Macaranga involucrata Baill. 
Lamarck, Encycl. 2 (1786) 206, referred it, with doubt, to Croton 
mauritianum Lam.; while Miquel, Fl. Ind. Bat. I 2 (1859) 385, 
thought that it might be Claoxylon indicum Hassk. 

ACALYPHA Linnaeus 

ACALYPHA AMENTACEA Roxb. Fl. Ind. ed. 2, 3 (1832) 676. 

Acalypha stipulacea Klotz. in Nov. Act. Acad. Nat. Cur. 19 (1843) 
Suppl. 1: 416. 

Acalypha amboynensis Benth. in Hook. Lond. Journ. Bot. 2 (1843) 233. 
Achyranthes spiciflora Burm. Index Alt. Herb. Amb. (1769) [5] 

(type!), non Acalypha spiciflora Burm. f. 
Cauda felis agrestis rubra Rumph. Herb. Amb. 4: 84, t. 37, f. 1. 
Cauda felis agrestis alba Rumph. Herb. Amb. 4: 84, t. '37, f. 2. 
Amboina, Robinson PL Rumph. Amb. 353, 354, July 22, 1913, along river 
banks at low altitudes in the vicinity of the town of Amboina, locally known 
as ekor kuching and ekor tusa. 

Apparently but a single species is represented by the two forms 
figured and described by Rumphius, and I consider both to be 
referable to Acalypha amentacea Roxb., the type of which was 
from the Moluccas. At the same time I do not see how the 
Philippine Acalypha stipulacea Klotz. can be distinguished speci- 
fically from the Amboina form, Klotzsch's species having been 
reported from as far to the southeast as New Guinea. Acalypha 
amboynensis Benth. and probably the Amboina specimen re- 
ferred by Bentham to Acalypha grandis Benth. doubtless are 
referable here. It is to be noted, however, that J. Mueller re- 
duced Acalypha amentacea Roxb. to Acalypha fruticosa Forsk., 
but the range of Forskal's species militates very greatly against 
the correctness of this reduction. Cauda felis agrestis rubra 
was reduced by Hasskarl, Neue Schlussel (1866) 79, to Acalypha 
grandis Benth., and no other disposition of it has been suggested. 
Cauda felis agrestis alba has been considered by numerous 
authors, but with little uniformity of opinion. Burman f., Fl. 
Ind. (1768) 203, erroneously placed it under Acalypha spiciflora 
Burm. f.=Cleidion spiciflorum (Burm. f.) comb. nov. (Cleidion 
javanicum Blume!). Linnaeus, Mant. 2 (1771) 127, erroneously 
placed it under Caturus spiciflorus Linn., which is a synonym of 
Acalypha hispida Burm., in which reduction he was followed by 
a few authors. Retzius, Obs. 5 (1789) 30, placed it under 
Acalypha betulina Retz., a species based on Ceylon specimens 
collected by Koenig and a synonym of Acalypha fruticosa Forsk. ; 



EUPHORBIACEAE 



323 



and finally Kosteletzsky, Allg. Med. Pharm. PL 5 (1836) 1743, 
placed it under Acalypha caturus Blume. 

ACALYPHA HiSPIDA Burm. f. Fl. Ind. (1768) 303, t. 61, f. 1. 
Acalypha densiflora Blume Bijdr. (1826) 628. 

Caturus spiciflorus Linn. Mant. 1 (1767) 127, non Acalypha spiciflora 
Burm. f. 

Cauda felis domestica Rumph. Herb. Amb. 4: 82, t. 36, 
Amboina, Robinson PL Rumph. Amb. 570, without data, apparently from 
cultivated plants. 

Burman f. described and figured Acalypha hispida from a 
Javan specimen, but reduced Cauda felis domestica Rumph. 
to his species which is certainly the correct disposition of it. 
Blume misinterpreted Burman's Acalypha hispida and rede- 
scribed the same form as Acalypha densiflora Blume, reducing 
to it also Cauda felis domestica Rumph. The plant is widely 
cultivated for ornamental purposes in the Malayan region. 

PLUKENETIA Linnaeus 

PLUKENETIA CORN ICU LATA Sm. in Nov. Act. Soc. Sci. Upsal. 4 
(1799) 4. 

Pterococcus glaberrimus Hassk. in Flora 25 (1842) Beibl. 41. 
Hedrayostylus glaberrimus Hassk. in Tijdschr. Nat. Wetensch. 10 
(1840) 141. 

' . Hedrayostylus corniculatus Hassk. Cat. Hort. Bot. Bogor. (1844) 234. 
Sajorium corniculatum Dietr. Syn. PI. 5 (1852) 331. 
Sajor volubilis Rumph. Herb. Amb. 1: 194, t. 79, f. 2. 

Not represented in our Amboina collections. Sajor volubilis 
Rumph. was originally reduced by Linnaeus to the American 
Plukenetia volubilis Linn., in Stickman Herb. Amb. (1754) 8, 
Amoen. Acad. 4 (1759) 120, Sp. PI. ed. 2 (1763) 1423, in which 
he was followed by Burman f ., Lamarck, Spanoghe, and Murray. 
The form figured and described by Rumphius has been cited 
under all or most of the various synonyms listed above. It was 
cited by Willdenow, Sp. PI. 4 (1805) 515, under Plukenetia cor- 
niculata Sm., and perhaps by Smith in the original description 
of the species, which I have not seen. Sajor volubilis Rumph. 
is doubtless correctly placed under Plukenetia corniculata Sm. 

RICINUS Linnaeus 

RICINUS COMMUNIS Linn. Sp. PI. (1753) 1007. 

Ricinus albus domesticus Rumph. Herb. Amb. 4: 92. 

Ricinus albus agrestis Rumph. Herb. Amb. 4: 92. 

Ricinus ruber Rumph. Herb. Amb. 4: 97, t. J>1. 
Amboina, Robinson PI. Rumph. Amb. 337, 338, in waste places about 
Castle Victoria, town of Amboina, November 13, 1913, locally known as 
jarak puti daun besaar and jar ok mera daun kechil. 



324 



RUMPHIUS'S HERBARIUM AMBOINENSE 



The Rumphian descriptions apply to three forms of this poly- 
morphous species. The reduction of Ricinus ruber was first 
made by Linnaeus, in Stickman Herb. Amb. (1754) 16, Amoen. 
Acad. 4 (1759) 126, Syst. ed. 10 (1759) 1276, Sp. PL ed. 2 
(1763) 1430. The various forms described by Rumphius have 
been reduced by authors to Ricinus lividus Jacq., R. viridis 
Willd., R. africanus Mill., and R. ruber Miq. FL Ind. Bat. I 2 
(1858) 390; the last is not included in Index Kewensis; all are 
synonyms of Ricinus communis Linn. 

ALE U RITES Forster 

ALEURITES MOLUCCANA (Linn.) Willd. Sp. PL 4 (1805) 590. 

Jatropha moluccana Linn. Sp. PL (1753) 1006. 

Croton moluccanus Lam. Encycl. 2 (1786) 207. 
] Aleurites triloba Forst. Char. Gen. (1776) 112. 

Juglans camirium Lour. Fl. Cochinch. (1790) 573. 

Aleurites ambinux Pers. Syn. 2 (1807) 587. 
I Camirium Rumph. Herb. Amb. 2: 180, t. 58. 

Amboina, Hitoe lama, Robinson PL Rumph. Amb. 332, October 8, 1913, 
cultivated, altitude about 175 meters, locally known as kamiri. 

Lamarck seems to have been the first author to reduce Cami- 
rium, placing it under Croton moluccanum Lam., which as 
Aleurites moluccana Willd. is certainly the correct disposition of 
it. By various authors it has been referred to Juglans camirium 
Lour., Aleurites triloba Forst., and A. ambinux Pers. — all syn- 
onyms of Aleurites moluccana (Linn.) Willd. 

JATROPHA Linnaeus 

JATROPHA CURCAS Linn. Sp. PL (1753) 1006. 

Ricinus americanus Rumph. Herb. Amb. 4: 95. 
Amboina, Binting, Robinson PI. Rumph. Amb. 333, September 16, 1913, 
along roadsides at low altitudes, locally known as jarap. 

This reduction was first made by Hasskarl, Neue Schliissel 
(1866) 81, and is certainly the correct disposition of the plant 
Rumphius described as Ricinus americanus. 

MAN I HOT Adanson 

MAN I HOT UTILISSIMA Pohl PL Bras. Ic. 1 (1827) 32, t. 2U. 

Jatropha manihot Linn. Sp. PL (1753) 1007. 

Yucca Rumph. Herb. Amb. 5: 325, 361. 

Mandihocca Rumph. Herb. Amb. 5: 361. 
Amboina, Soja, Robinson PI. Rumph. Amb. 336, a roadside escape, 
altitude about 375 meters. 

The reductions follow Hasskarl, Neue Schliissel (1866) 128, 
133, and are unquestionably correct. 



EUPHORBIACEAE 



325 



CODIAEUM * Jussieu 

CODIAEUM VARI EGATUM (Linn.) Blume Bijdr. (1825) 606. 

Croton variegatus Linn. Sp. PI. (1753) 1199 (type!). 

Codiaeum simplex (chrysosticton ) Rumph. Herb. Amb. 4: 65, t. 25. 
Amboina. Wae and Soja, Robinson PL Rumph. Amb. 3^8, 3^9, October 
and November, 1913, cultivated and on coral hillsides at low altitudes, 
locally known as kodiho. Additional specimens, perhaps referable to one 
or the other of the numerous forms described by Rumphius are Robinson 
PI. Rumph. Amb. 3J+3, 3UU, 3^5, from cultivated plants, Soja. October 24, 
1913, all sterile. 

Croton variegatus Linn, is one of the few species published in 
the first edition of the Species Plantarum that were based entirely 
on Rumphius. It is typified by Codiaeum chrysosticton Rumph., 
Herb. Amb. 4: 65, t. 25. Here Linnaeus also reduced, as varie- 
ties, Codiaeum taeniosum Rumph., 1. c. 68, t. 26, and Codiaeum 
silvestre Rumph., 1. c. 69, t. 27, the latter, however, belonging 
with Codiaeum bractiferum Roxb. These reductions were re- 
peated in the later writings of Linnaeus, Stickman Herb. Amb. 
(1754) 16, Amoen. Acad. 4 (1759) 126, Syst. ed. 10 (1759) 
1275, Sp. PL ed. 2 (1763) 1424, and were generally accepted by 
later authors. Hasskarl, Neue Schliissel (1866) 76, has reduced 
many of the numerous forms characterized by Rumphius to 
varieties of Codiaeum variegatum (Linn.) Blume, but the distinc- 
tions between the very numerous varieties and forms of this 
polymorphous species are vague and unsatisfactory. The forms 
characterized by Rumphius are as follows : 

Codiaeum chrysosticton medium. 
Codiaeum chrysosticton latifolium. 
Codiaeum chrysosticton angustifolium 
Codiaeum chrysosticton medium rubrum. 
Codiaeum chrysosticton rubro-maculatum. 
Codiaeum erythrosticton parvifolium. 
Codiaeum nigrum medium. 
Codiaeum nigrum minus. 
Codiaeum parvifolium viride. 

CODIAEUM VARIEGATUM (Linn.) Blume f. TAENIOSUM Muell.-Arg. in 

DC. Prodr. 15 2 (1866) 1120. 
Codiaeum taeniosum Rumph. Herb. Amb. 4: 68, t. 26, f. 1. 

Amboina, Soja, Robinson PL Rumph. Amb. 346, 3U7, October 24, 1913, 
cultivated, altitude about 300 meters. 

A narrow-leaved form of Codiaeum variegatum (Linn.) Blume, 
originally placed by Linnaeus as var. p Sp. PI. (1753) 1199. 
Rumphius describes two forms under Codiaeum taeniosum, lu- 

* Retained name, Vienna Code; Phyllaurea Lour. (1790) is older. 



326 



RUMPHIUS'S HERBARIUM AMBOINENSE 



teum and viride, the latter, t. 26, f. 2, with crisped leaves= Croton 
variegatum var. crispum Muell.-Arg. 

CODIAEUM BRACTI FERU M Roxb. Fl. Ind. ed. 2, 3 (1832) 680. 

Codiaeum brevistylum Pax & K. Hoffm. in Engl. Pflanzenreich 47 
(1911) 28. 

Codiaeum silvestre Rumph. Herb. Amb. 4: 69, t. 27. 
Amboina, Koesoe koesoe sereh, Robinson PI. Rumph. Amb. 3J/.0, August 
23, 1913, with staminate and pistillate flowers. Probably referable here 
are Robinson PI. Rumph. Amb. 341, 342, from coral hillsides at Wae, 
November 28, 1913, both sterile, and one lacking the characteristic reduced 
"bracts" figured by Rumphius and described by Roxburgh. 

Codiaeum silvestre Rumph. was originally reduced by Linnaeus 
to Codiaeum variegatum Linn. var. y Sp. PL (1753) 1199, but 
represents an entirely distinct species, readily separated by its 
pubescent ovaries and its vegetative characters. The greatly 
reduced, bract-like leaf is very characteristic, but is present 
opposite only the uppermost leaves and apparently falls early. 
Codiaeum bractiferum Roxb. is not listed in Index Kewensis 
and was described as follows : 

C. bractiferum. R. Shrubby. Leaves linear-oblong, smooth, entire. 
Racemes terminal, becoming lateral with an immense orbicular bract at 
the base. Codiaeum silvestre Rumph. Amb. IV. t. 27. A native of the 
Moluccas. 

I have placed the recently described Amboinese Codiaeum 
brevistylum Pax & K. Hoffm. as a synonym, although in the de- 
scription the characteristic, reduced, bract-like leaves are not 
mentioned, possibly because they had fallen from the specimens 
collected by Dolleschal. 

ENDOSPERMUM Bentham 

EN DOSPERM U M MOLUCCANUM (Teysm. & Binn.) Becc. Malesia 2 
(1884) 38. 

Capellania moluccana Teysm. & Binn. in Nat. Tijdschr. Ned. Ind. 29 
(1866) 239. 

Arbor regis Rumph. Herb. Amb. 2: 257, t. 85. 

This species is not represented in our Amboina collections; 
however, it has been collected in Amboina by Dolleschal. Lin- 
naeus originally referred Arbor regis Rumph., with doubt, to 
Hernandia sonora Linn., in Stickman Herb. Amb. (1754) 11, 
Amoen. Acad. 4 (1759) 122, thus: "85. Arbor regis= Hernandia 
sonora ? (sed fructus alienus.)." In this reduction he was 
followed by Burman f ., Lamarck, Willdenow, Persoon, Henschel, 
Spanoghe, Dietrich, Miquel, and other authors. Beccari, Ma- 
lesia 2 (1884) 38, first made the correct reduction of it to 



EUP HORBI ACEAE 



327 



Endospermum moluccanum (Teysm. & Binn.) Becc, the type 
of Capellania moluccana Teysm. & Binn. being from the Mo- 
luccas. Teysmann, quoted by Hasskarl, Neue Schlussel (1866) 
45, thought that it was referable to Tournefortia. 

EXCOECARIA Linnaeus 

EXCOECARIA AGALLOCHA Linn. Syst. ed 10 (1759) 1288, Amoen. Acad. 
4 (1759) 122, Sp. PL ed. 2 (1763) 1451 (type!). 
Arbor excoecans Rumph. Herb. Amb. 2: 237, t. 79 <$, t. 80 ?. 
Amboina, Paso, Robinson PL Rumph. Amb. 385, October 29, 1913, along 
the seashore, locally known as kayu matta buta. 

The Rumphian illustration and description are the whole 
basis of Excoecaria agallocha Linn, as originally published by 
Linnaeus. The reduction, certainly correct, has been accepted 
by all authors. The form briefly described by Rumphius as 
Arbor excoecans (II variegata), Herb. Amb. 2: 239, can scarcely be 
other than this species. 

HOMALANTHUS Jussieu 

HOM ALANTH US POPULNEUS (Geisel.) Pax in Engl. & Prantl Nat. 
Pflanzenfam. 3 6 (1890) 96, /. 60. 

Stillingia populnea Geisel. Croton. Monogr. (1807) 80. 
Carumbium populneum Muell.-Arg. in DC. Prodr. 15 2 (1866) 1144. 
Frutex excoecans Rumph. Herb. Amb. 4: 130, t. 65. 
Amboina, Koesoekoesoe sereh, Robinson PI. Rumph. Amb. 329, October 3, 
1913, in light forests, locally known as daun matta bali. 

Teysmann, quoted by Hasskarl, Neue Schlussel (1866) 85, 
reduced Frutex excoecans Rumph. to Carumbium populifolium 
Reinw., which is a synonym of Homalanthu^ populneus (Geisel.) 
Pax. The Amboina form appears to be typical Homalanthus 
populneus Pax [that is, var. genuinus Pax in Engl. Pflanzenreich 
52 (1912) 46], which extends from the Malay Peninsula to Java, 
Sumatra, Borneo, and Celebes. 

PIMELEODENDRON Hasskarl 

PIMELEODENDRON AMBOINICUM Hassk. Versl. en Med. Akad. Ams- 
terdam 4 (1855) 140. 
Carumbium amboinicum Miq. Fl. Ind. Bat. 1 2 (1859) 413. 
Arbor pinguis Rumph. Herb. Amb. 2: 249, t. 83. 
Amboina, Hoenoet, Robinson PI. Rumph. Amb. 174, October 13, 1913, at 
low altitudes, locally known as mamina. 

Arbor pinguis Rumph. was first correctly placed by Hasskarl 
as a synonym of Pimeleodendron amboinicum Hassk. Desrouss, 
in Lam. Encycl. 3 (1791) 693, thought that it might belong in 



328 



RUMPHIUS'S HERBARIUM AMBOINENSE 



the Sapotaceae, while Henschel, Vita Rumph. (1833) 152, thought 
that it might be Cambogia gutta Linn. The species is known 
only from Amboina. 

SAPIUM P. Browne 

SAPIUM INDICUM Willd. Sp. PI. 4 (1805) 572. 

Ichthyoctonos litorea Rumph. Herb. Amb. 3: 213, t. 138. 

This species is not represented in our Amboina collections, 
but I reduce the Rumphian species to Sapium indicum Willd. 
with great confidence that this is the correct disposition of it. 
Baillon, Etud. Gen. Euphorb. (1858) 518, suggested that it 
was a Stillingia; while Hasskarl, Neue Schlussel (1866) 68, 
reduced it to Excoecaria virgata Zoll. & Mor. =Sapium virgatum 
Hook, f., a species very closely allied to Sapium indicum Willd. 
Sapium virgatum Hook. f. is confined to Java, according to 
Pax & K. Hoffmann, in Engl. Pflanzenreich 52 (1912) 250, while 
Sapium indicum Willd. extends from India to New Guinea. The 
figure given by Rumphius is fairly good, while the description 
applies to Sapium indicum Willd. in all respects. 

EUPHORBIA Linnaeus 

EUPHORBIA HIRTA Linn. Sp. PL (1753) 454. 

Euphorbia pilulifera Linn. Sp. PI. (1753) 454. 

Euphorbia capitata Lam. Encycl. 2 (1788) 422. 

Esula esculenta Rumph. Herb. Amb. 6: 54, t. 23, f. 2. 
Amboina, Hitoe lama, Robinson PI. Rumph. Amb. 352, October 8, 1913, 
in cultivated ground, altitude about 150 meters; also represented by Rel. 
Robins. 2497, from Boeton Island, July 13, 1913. 

Esula esculenta Rumph. was first reduced by Linnaeus to 
Euphorbia hirta Linn., in Stickman Herb. Amb. (1754) 26, 
Amoen. Acad. 4 (1759) 134, Syst. ed. 10 (1759) 1048, Sp. PL 
ed. 2 (1762) 651, which is certainly the correct disposition of 
it, and in which he was followed by many authors. In recent 
literature the species appears chiefly as Euphorbia pilulifera 
Linn., which is a synonym of Euphorbia hirta Linn., the latter 
having page priority. Lamarck referred it, with doubt, to 
Euphorbia capitata Lam., a synonym of Euphorbia hirta Linn. 

EUPHORBIA N ERI I FOLIA Linn. Sp. PI. (1753) 451. 

Euphorbia ligularia Roxb. Hort. Beng. (1814) 36, nomen nudum, FL 

Ind. ed. 2, 2 (1832) 465. 
Ligularia lactea Rumph. Herb. Amb. 4: 88, t. 40. 
Ligularia minor Rumph. Herb. Amb. 4: 90. 

This species is not represented in our Amboina collections. 
This reduction of Ligularia lactea Rumph. was first made by 



ANACARDIACEAE 



329 



Linnaeus, in Stickman Herb. Amb. (1754) 16, Amoen. Acad. 
4 (1759) 126, Sp. PI. ed. 2 (1762) 648, and is apparently the 
correct disposition of it. Euphorbia ligularia Roxb. was de- 
scribed from Bengal specimens, to which Ligularia Rumph. 
was reduced. Hasskarl, Neue Schliissel (1866) 81, referred 
Ligularia minor Rumph. to Euphorbia edulis Lour., but the form 
described by Rumphius was probably only a somewhat reduced 
one of Euphorbia nerii folia. 

Ligularia lactea e Java, briefly described by Rumphius, Herb. 
Amb. 4:88, was thought by Hasskarl, Neue Schliissel (1866) 
80, to be possibly the same as Euphorbia splendens Boj., but 
this reduction is probably incorrect. 

EUPHORBIA TIRUCALLI Linn. Sp. PI. (1753) 452. 
Ossifraga lactea Rumph. Herb. Amb. 7: 62, t. 29. 

This species is not represented in our Amboina collections. 
Ossifraga lactea was first reduced to Euphorbia tirucalli Linn, 
by Linnaeus, in Amoen. Acad. 4 (1759) 137, Syst. ed. 10 (1759) 
1047, Sp. PL ed. 2 (1762) 649. This is manifestly the correct 
disposition of it and has been accepted by all authors who have 
had occasion to cite the Rumphian illustration. 

EUPHORBI ACEAE indet. 

Involucrum cusci Rumph. Herb. Amb. 4: 73. 

The description, while rather short, is sufficiently definite to 
indicate that a euphorbiaceous plant is intended. I am, however, 
unable to refer it to its proper genus from the description alone. 
Hasskarl, Neue Schliissel (1866) 77, suspects it to be an 
Elateriospermum or a species of some allied genus. 

ANACARDIACEAE 

MANGIFERA Linnaeus 

MANGIFERA FOETIDA Lour. Fl. Cochinch. (1790) 160. 
Manga foetida Rumph. Herb. Amb. 1: 98, t. 28. 
Amboina, Hitoe lama, Robinson PI. Rumph. Amb. 126, beside clearings, 
altitude about 175 meters, locally known as ambachan, bachan, and bichang. 

Manga foetida Rumph. was reduced by Loureiro in the original 
description of Mangifera foetida Lour, and is undoubtedly the 
form that Loureiro described from Cochin-China material, as 
the species is widely distributed in the Indo-Malayan region 
in cultivation. All authors have followed Loureiro in this 
reduction. 



330 



RUMPHIUS'S HERBARIUM AMBOINENSE 



MANGIFERA CAESIA Jack in Roxb. Fl. Ind. 2 (1824) 441. 
Mangifera kemanga Blume Mus. Bot. 1 (1850) 202. 
Manga foetida II Rumph. Herb. Amb. 1: 99. 

This species is not represented in our Amboina collections. 
The reduction follows Blume, who cites Mangifera foetida II 
under the native name wani in the original description of Man- 
gifera kemanga. It should be also compared with Mangifera 
odorata Griff. 

MANGIFERA UTANA Ham. in Mem. Wern. Soc. 5 2 (1826) 326 (type!). 

Mangifera membranacea Blume Mus. Bot. 1 (1850) 195 (type?). 
Mangifera taipan Ham. ex Miq. Fl. Ind. Bat. 1 2 (1858) 631. 
Manga silvestris I Rumph. Herb. Amb. 1: 97, t. 27. 

Nothing resembling this species occurs in our Amboina col- 
lections. The Rumphian figure and description are the whole 
basis of Mangifera utana Ham., a species properly published, 
but overlooked by the compilers of Index Kewensis, in which 
it is not listed. Mangifera membranacea Blume is based, at 
least in part, on the same Rumphian description and figure ; like 
Mangifera utana Ham. it is a species of very doubtful status, 
placed by Engler, in DC. Monog. Phan. 4 (1883) 215, under the 
■ heading "species omnino incertae." The figure very closely 
resembles the Philippine form, Mangifera monandra Merr. 

MANGIFERA INDICA Linn. Sp. PL (1753) 200. 

Manga domestica Rumph. Herb. Amb. 1: 93, t. 25, 26. 

Amboina, Paso, Robinson PI. Rumph. Amb. 125, October 29, 1913, locally 
known as manga; town of Amboina, PI. Rumph. Amb. 123, 124., August 24, 
1913, locally known as pan and manga pan. 

Linnaeus originally reduced both tt. 25 and 26 to Mangifera 
indica Linn., in Stickman Herb. Amb. (1754) 7, Amoen. Acad. 
4 (1759) 119; and practically all authors have followed him, 
at least in the reduction of Manga domestica as represented by 
t. 25. The form represented by t. 26 is almost certainly nothing 
but Mangifera indica; at least no characters are indicated by 
which it can be distinguished. It has been cited by Blume and 
by Miquel as Mangifera altissima Blanco, a very characteristic 
species that has very little in common with the figure given by 
Rumphius ; whatever else t. 26 may represent, it is certainly not 
Mangifera altissima Blanco. 

Rumphius describes five forms of Manga domestica, all of 
which, with one exception, are surely referable to Mangifera 
indica Linn. The exception is Arbor mangifera V minor, which 
Blume has reduced to Mangifera minor Blume, Mus. Bot. 1 



ANACARDIACEAE 



331 



(1850) 198, his description being based on specimens from 
Celebes. 

MANGIFERA MINOR Blume Mus. Bot. 1 (1850) 198. 

Manga domestica minor Rumph. Herb. Amb. 1: 94. 

Mangifera minor Blume was based on specimens from Celebes, 
and Blume reduced the Rumphian species as a synonym in the 
original description of the species. The species is a valid one, 
placed by Engler near Mangifera longipes Griff.; DC. Monog. 
Phan. 4 (1883) 202. Blume was undoubtedly correct in re- 
ducing here Manga minor Rumph. 

MANGIFERA LAURINA Blume Mus. Bot. 1 (1850) 195. 
Manga simiarum Rumph. Herb. Amb. 1: 94. 

The reduction follows Blume, who placed the Rumphian species 
as a synonym of Mangifera laurina Blume in the original de- 
scription of the species. 

MANGIFERA TAI PAN Ham. in Mem. Wern. Soc. 5 2 (1826) 326 (type!). 
Mangifera silvestris altera Rumph. Herb. Amb. 1: 97. 

This species is based wholly on Rumphius's description and 
is of entirely doubtful status. Engler, in DC. Monog. Phan. 4 
(1883) 215, reduces it to Mangifera membranacea Blume, Mus. 
Bot. 1 (1850) 195, following Blume, but Blume's species is in 
turn one of entirely doubtful status. Hamilton's name is the 
older and should be maintained if future investigations show 
that the species is ? valid one. 

MANGIFERA RUMPHII Pierre Fl. Forest. Cochinch. 4 (1897) t. S6U, f- E, 
excl. syn. Blanco. 

Pauw Rumph. Herb. Amb. 7: 18, t. 11 (incl. / maxima, II media, 
III minima). 

The three forms described by Rumphius under the name 
Pauw, were referred by Hasskarl to Mangifera altissima Blanco, 
but the descriptions and the figure pertain to a totally different 
species, which Pierre has described from Banda specimens as 
Mangifera rumphii Pierre. 

GLUT A Linnaeus 
GLUTA BENGHAS Linn. Mant. 2 (1771) 293. 

Terminalia vernix Lam. Encycl. 1 (1783) 350 (type!). 
Stagmaria verniciflua Jack. Malay Miscel. 3 (1823) 12. 
Arbor vernicis Rumph. Herb. Amb. 2: 259, t. 86. 

Gluta benghas Linn, was based on Javan specimens, perhaps 
benghas being a typographical error, as the specific name is 
taken from its native name rengas. Arbor vernicis Rumph. 



332 RUMPHIUS'S HERBARIUM AMBOINENSE 

probably includes more than Gluta benghas Linn., but the figure 
and the description, at least for the most part, apparently belong 
here. Loureiro, Fl. Cochinch. (1790) 587, discusses it under 
Vernicia montana Lour., but does not refer it to this species, 
which is supposed to be Aleurites moluccana Willd. It is the 
type and whole basis of Terminalia vernix Lam. Hasskarl seems 
to have been the first author to reduce it to Gluta benghas Linn., 
Flora (1844) 619, in which he was followed by Endlicher, 
Blume, and Miquel, and, very generally, by recent authors. 
Miquel also discusses it following Terminalia angustifolia Jacq., 
Fl. Ind. Bat. I 1 (1855) 599, overlooking the fact that Terminalia 
vernix Lam., which he considered to be a doubtful species, was 
based on Arbor vernicis Rumph., which in the same work, l 2 
(1858) 624, he correctly placed under Gluta benghas Linn. 

SPONDIAS Linnaeus 

SPONDIAS DULCIS Forst. Prodr. (1786) 34. 
Evia acida Blume Mus. Bot. 1 (1850) 234. 

Spondias acida Blume ex Steud. Nomencl. ed. 2, 2 (1841) 625. 
Spondias dulcis Forst. var. acida Engl, in DC. Monog. Phan. 4 (1883) 
247. 

Condondum Rumph. Herb. Amb. 1: 161, t. 60. 

This is not represented in our Amboina collections, but the 
species is known to occur in the island. I have followed Blume 
in this reduction, as it is very probable that he was correct in 
referring Condondum Rumph. to Evia acida Blume. Engler 
considers that Blume's Evia acida is a variety of Spondias dulcis 
Forst. Linnaeus originally reduced Condondum to Chrysobala- 
nus icaco Linn., in Stickman Herb. Amb. (1754) 8, Amoen. Acad. 
4 (1759) 119, a pure error on his part, the reduction being 
abandoned in his later writings. Lamarck, Encycl. 3 (1791) 
697, followed by Hamilton, Mem. Wern. Soc. 5 2 (1826) 358, 
thought that it was Mangifera pinnata Linn, f., which was 
figured by Rumphius under the name Condondum malaccense, 
t. 61. Mangifera pinnata Lam., non Linn, f., is Sorindeia 
madagascariensis Thouars. 

SPONDIAS PINNATA (Linn, f.) Kurz in Pegu Report (1875) A 44. 
Mangifera pinnata Linn. f. Suppl. (1781) 156. 
Evia amara Commers. ex Blume Mus. Bot. 1 (1850) 234. 
Spondias mangifera Willd. Sp. PI. 2 (1799) 751. 
Spondias amara Lam. Encycl. 4 (1797) 261. 
Condondum malaccense Rumph. Herb. Amb. 1: 162, t. 61. 

This is not represented in our Amboina collections. The 
reduction, apparently the correct disposition of Condondum 



ANACARDIACEAE 



333 



malaccense Rumph., follows Blume and other authors. Blume 
makes the Rumphian figure and description the type of Evia 
amara Commers. var. tuberculosa Blume; and Engler, recogniz- 
ing the variety, transfers it to Spondias mangifera Willd. var. 
tuberculosa Blume. The form described and figured, however, 
is probably nothing but typical Spondias pinnata (Linn, f.) 
Kurz. 

DRACONTOMELUM Blume 

DRACONTOMELUM M ANGI FERUM Blume Mus. Bot. 1 (1850) 231. 
Poupartia mangifera Blume Bijdr. (1826) 1160, excl. syn. 
Pomum draconum Rumph. Herb. Amb. 1: 157, t. 58. 
Amboina, Kaju poeti, Robinson PL Rumph. Amb. 127, October 24, 1913, 
in light woods, altitude about 200 meters, locally known as bua rau. 

Hamilton, Mem. Wern. Soc. 5 2 (1828) 358, thought Pomum 
draconum Rumph. to be "pretty clearly a Spondias," but Blume, 
Mus. Bot. 1 (1850) 231, referred it to Dracontomelum mangi- 
ferum Blume, .which is apparently the correct disposition of it, 
and one that has been accepted by subsequent authors, including 
Walpers, Miquel, and Engler. The species is widely distributed 
in the Malay Archipelago. 

DRACONTOMELUM SYLVESTRE Blume Mus. Bot. 1 (1850) 231. 

Pomum draconum silvestre Rumph. Herb. Amb. 1: 159, t. 59. 

This is not represented in our Amboina collections. The 
reduction was made by Blume in the original description of 
Dracontomelum sylvestre and is probably the correct disposition 
of it. According to Blume the species is widely distributed in 
the Malay Archipelago, but in the most recent monograph of the 
group Engler cites specimens from Borneo only. 

AN ACARDI U M Linnaeus 
AN ACARDI U M OCCI DENTALE Linn. Sp. PL (1753) 383. 
Cassuvium pomiferum Lam. Encycl. 1 (1783) 22. 
Cassuvium Rumph. Herb. Amb. 1: 177, t. 69. 

The common cashew is not represented in our Amboina col- 
lections, but is very generally cultivated throughout the Malayan 
region. Cassuvium Rumph. was originally reduced by Linnaeus 
to Anacardium occidentale Linn., in Stickman Herb. Amb. 
(1754) 8, Amoen. Acad. 4 (1759) 120, Syst. ed. 10 (1759) 1019, 
Sp. PL ed. 2 (1762) 548, which reduction is certainly correct 
and has been very generally followed by subsequent authors. 
Lamarck, however, redescribed the same species as Cassuvium 
pomiferum Lam., referring here Cassuvium Rumph., from which 
he took the generic name. 



334 



RUMPHIUS'S HERBARIUM AMBOINENSE 



SEMECARPUS Linnaeus f. 

SEMECARPUS CASSUVIUM Roxb. Hort. Beng. (1814) 32 (type!); 
Spreng. Syst. 1 (1825) 936; Roxb. Fl. Ind. ed. 2, 2 (1832) 85. 

Anacardium longifolium Lam. Encycl. 1 (1763) 139, p. p., quoad syn. 
Rumph. 

Cassuvium silvestre Rumph. Herb. Amb. 1: 179, t. 70. 

Amboina, Paso, Way tommo, and near the town of Amboina, Robinson 
PL Rumph. Amb. 119, 120, 121, 122, October, November, 1913, in thickets 
and light woods, sea level to an altitude of 20 meters, locally known as saku. 

Cassuvium silvestre Rumph. is the whole basis of Semecarpus 
cassuvium Roxb. as originally published by him, Hort. Beng. 
(1814) 32, by citation.* Sprengel apparently based his short 
description partly, if not entirely, on Anacardium longifolium 
Lam., which is the same as Semecarpus cassuvium Roxb. only 
in small part. Roxburgh's actual description, as published in 
1832, was based on specimens originating in the Moluccas and 
cultivated in the botanic garden at Calcutta. 

SEMECARPUS FORSTEN 1 1 Blume Mus. Bot. 1 (1850) 188. 

Cassuvium silvestre s. Lau Lassi (e Ternate) Rumph. Herb. Amb. 
1: 180. 

The reduction follows Blume, who so reduced the Ternate 
form mentioned by Rumphius, in the original description of 
Semecarpus forstenii Blume, which was based on actual speci- 
mens collected in Ternate by Forsten. 

CELASTRACEAE 

EUONYMUS Linnaeus 

EUONYMUS sp.? 

Caju lape lape Rumph. Herb. Amb. 3: 78, t. 50. 

Nothing resembling the form figured by Rumphius occurs in 
our Amboina collections. The figure looks suspiciously like 
certain species of Euonymus, while the description applies fairly 
closely. Blume, Rumphia 3 (1837) 167, notes that it cannot be 
a M is cho carpus ; Miquel, Fl. Ind. Bat. 1 2 (1858) 567, mentions 
it under Cupania fuscescens M.iq.=Mischocarpus fuscescens 
Blume; while Hasskarl, Neue Schliissel (1866) 53, quotes Teys- 
mann's opinion that it was an Elodea (Tridesmis) =Cratoxylon, 
which is an impossible disposition of it. It may possibly be a 
sapindaceous plant, but the probabilities are that it is a poorly 
described and figured species of Euonymus. 



* See C. B. Robinson in Philip. Journ. Sci. 7 (1912) Bot. 413, 418. 



ICACINACEAE 



335 



PERROTTETI A Humbolt, Bonpland, and Kunth 

PERROTTETIA MOLUCCANA (Blume) Loesen. in Engl. & Prantl. Nat. 
Pflanzenfam. 3 5 (1892) 220. 
Caryospermum moluccanum Blume Mus. Bot. 1 (1850) 176. 
Vertifolia rubra Rumph. Herb. Amb. 3: 100, t. 67. 
Amboina, Hitoe messen, Mahija, and Lateri, Robinson PL Rumph. Amb. 
157, 160, 601, August and November, 1913, in forests, altitude 100 to 250 
meters. 

No previous reduction of Vertifolia rubra Rumph. has been 
suggested other than Hasskarl's tentative reference of it to the 
Euphorbiaceae. The specimens cited above agree perfectly with 
the description and very well with the figure, the fault in the 
figure being that the flowers are relatively greatly enlarged. 
Hasskarl was uncertain to which description, Vertifolia alba or 
Vertifolia rubra, the figure pertained, as Rumphius does not 
indicate which one he intended it for. The descriptions pertain 
to two entirely different plants, of different genera, and probably 
of different families. The Amboina material shows definitely 
that the figure belongs with Vertifolia rubra Rumph. It is of 
interest to note that the species is the type of the genus Caryo- 
spermum and that Blume's material was from Amboina. 

ICACINACEAE 

CARDIOPTERYX Wallich 
CARDIOPTERYX MOLUCCANA Blume Rumphia 3 (1847) 207. 
Olus sanguinis Rumph. Herb. Amb. 5: 482, t. 180. 

This is not represented in our Amboina collections. Olus 
sanguinis Rumph. was originally reduced by Linnaeus, through 
error, to Dioscorea sativa Linn., in Stickman Herb. Amb. (1754) 
25, Amoen. Acad. 4 (1759) 133, in which he was followed by a 
number of authors. It has also been referred to Dioscorea clif- 
fortiana Lam. and to D. deltoidea Wall. Blume placed it in 
Cardiopteryx (often spelled Cardiopteris) in the original descrip- 
tion of Cardiopteryx moluecana Rlume, which is manifestly the 
correct disposition of it. 

STEMONURUS Blume 

STEMONURUS sp. 

Fructus bobae Rumph. Herb. Amb. 3: 166, 1. 105. 

Nothing resembling this form is represented in our Amboina 
collections. The figure is an excellent one and presents an icacin- 
aceous plant, undoubtedly of the genus Stemonurus. Teysmann, 
quoted by Hasskarl, Neue Schliissel (1866) 62, placed it in Platea, 
another genus of the same family. 



336 



RUMPHIUS'S HERBARIUM AMBOINENSE 



SAPINDACEAE * 

CARDIOSPERMUM Linnaeus 

CARDIOSPERMUM H ALICACABU M Linn. Sp. PL (1753) 366. 

Halicacabus peregrinus Rumph. Herb. Amb. 6: 61, t. 2J>, f. 2. 

This common and well-known species is not represented in our 
Amboina collections. The reduction of the Rumphian Halica- 
cabus 'peregrinus was first made by Linnaeus, in Stickman Herb. 
Amb. (1754) 26, Amoen. Acad. 4 (1759) 134, Syst. ed. 10 (1759) 
1007, Sp. PL ed. 2 (1762) 525, which has been followed by all 
succeeding authors. The exact form figured by Rumphius is 
apparently Cardiospermum halicacabum Linn. var. microcarpum 
Blume. 

ALLOPHYLUS Linnaeus 

ALLOPHYLUS TIMORENSIS (DC.) Blume Rumphia 3 (1847) 130 (emend. 
Radlk.). 

Schmidelia timorensis DC. Prodr. 1 (1824) 611. 

Ampacus litorea prima Rumph. Herb. Amb. 2: 188. 
Amboina, near the town of Amboina, Robinson PL Rumph. Amb. 53, 
August 8, 1913, along the beach, with the doubtful native names biking ubat 
and kaya besi. 

The position of this form, as described by Rumphius, has 
not previously been indicated, but Ampacus litorea prima is 
unquestionably identical with Allophylus timorensis (DC.) 
Blume. 

ALLOPHYLUS TERN AT US (Forst.) Radlk. in Engl. & Prantl Nat. Pflan- 
zenfam. 3 5 (1895) 313. 
Aporetica ternata Forst. Char. Gen. (1776) 132. 
Allophylus amboinensis Blume Rumphia 3 (1847) 129. 
Ampacus litorea (angustifolia) minor Rumph. Herb. Amb. 2: 189. 
Amboina, Amahoesoe, Robinson PL Rumph. Amb. 52, September 18, 
1913, along the beach. Another form of the same species is represented 
by Rel. Robins. 1603, from Paso, Amboina, Oct. 29, 1913, with elongated 
racemes, approaching Allophylus leptococcus Blume. 

Blume, Rumphia 3 (1847) 129, referred Ampacus litorea 
(angustifolia) minor to his Allophylus amboinensis, which is 
an exact synonym of Allophylus ternatus (Forst.) -Radlk. 
Miquel, Fl. Ind. Bat. I 2 (1859) 575, referred it to Allophylus 
timorensis (DC.) Blume, but Ampacus litorea prima Rumph. is 
manifestly Allophylus timorensis Blume, while the form from 
Leytimor is surely A. ternatus Radlk. (A. amboinensis Blume). 

It is to be noted that Ampacus angustifolius Rumph., Herb. 



* I am indebted to Dr. L. Radlkofer for the identifications of the speci- 
mens cited in this family. 



SAPINDACEAE 



337 



Amb. 2 : 188, t. 62, as described and figured, under which the two 
species discussed above are briefly described, is no Allophylus, 
but is Evodia amboinensis Merr. (Evodia triphylla auct. plur., 
non DC.) (see p. 290). 

SAPINDUS Linnaeus 

SAPINDUS RARAK DC. Prodr. 1 (1824) 608. 

Dittelasma rarak Hook. f. in Benth. & Hook. f. Gen. PI. 1 (1862) 395. 
Saponaria Rumph. Herb. Amb. 2: 134. 

This is not represented in our Amboina collections, but Sapon- 
aria is unquestionably the same as Sapindus rarak DC, which 
was based in part on Rumphius and in part on a Javan specimen. 
Rumphius notes that the plant was abundant in Java, but oc- 
curred in Amboina only as an introduced one. Burman f., Fl. 
Ind. (1768) 91, referred it to Sapindus saponaria Linn., in which 
he was followed by Loureiro, Fl. Cochinch. (1790) 238, but 
Sapindus saponaria Linn, is a species entirely distinct from S. 
rarak DC. 

LEPISANTHES Blume 

LEPISANTHES sp.? 

Arbor palorum alba parvifolia Rumph. Herb. Amb. 3: 98, t. 65 
(excl. /. A.). j ! 

The form figured and described by Rumphius is of doubtful 
status. Poiret, in Lam. Encycl. Suppl. 3 (1813) 479, thought 
that it might be Pometia pinnata Forst., which is certainly 
incorrect. De Candolle, Prodr. 1 (1824) 615, erroneously 
identified it as Stadmannia sideroxylon DC. Blume, Rumphia 3 
(1837-47) 149, thought that it might be the same as Scoro- 
dendron pollens Blume= Lepisanthes pollens Radlk., which was 
described from Timor specimens. Teysmann, quoted by Hass- 
karl, Neue Schliissel (1866) 55, placed it under Irine=Mna 
=Pometia, where it certainly does not belong. If Blume's 
surmise that it is identical with Scorodendron pollens is correct, 
it is strange that Rumphius did not mention the peculiar onion- 
like odor that is characteristic of Blume's species. 

SCH LEICH ERA Willdenow 

SCH LEICH ERA OLEOSA (Lour.) comb. nov. 

Pistacia oleosa Lour. Fl. Cochinch. (1790) 615. 
Schleichera trijuga Willd. Sp. PI. 4 (1805) 1096. 
Cussambium spinosum Ham. ex Henschel Vita Rumph. (1833) 143 
(type!). 

Cussambium Rumph. Herb. Amb. 1: 154, t. 57. 

This species is not represented in our Amboina collections, but 

144971 22 



338 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Cussambium, although poorly figured by Rumphius, is certainly 
referable to this widely distributed Indo-Malayan species. It 
was first reduced by Loureiro, Fl. Cochinch. (1790) 615, to his 
Pistacia oleosa, and the plant he described from specimens cul- 
tivated in Cochin-China is apparently identical with Schleichera 
trijuga Willd. ; at any rate it is no Pistacia. Accordingly, 
Loureiro's specific name, being the older, is here adopted. The 
reference of Cussambium to Schleichera trijuga Willd. was first 
made by Blume, Rumphia 3 (1847) 147, who also cites Pistacia 
oleosa Lour, as a synonym. Cussambium spinosum Ham. is 
typified by the Rumphian plate and description. 

EUPHORIA Commerson 

EUPHORIA LONGANA Lam. Encycl. 3 (1791) 574. 

Nephelium long ana Cambess. in Mem. M >• > 8 1 *• o :>•> 
Linkeng Rumph. Herb. Amb. 1 : 157. 

This Chinese species is briefly mentioned by Rumphius, who 
discusses it under the Chinese name linkeng. The material he 
had was unquestionably Euphoria longana Lam. 

CUBILIA Blume 

CUBILIA BLANCOI Blume Rumphia 2 (1849) 101. 
Euphoria cubili Blanco Fl. Filip. (1837) 287. 

Cubilia rumphii Blume 1. c. 101; Koord. in Ic. Bogor. 1 (1904) 51, 

t. 92, 93. 

Boa massy Rumph. Herb. Amb. 7: 5, t. 3. 

This species is not represented in our Amboina collections, 
but is now known from a number of localities in the Philippines 
(Luzon to Mindanao), from Bum, and from Celebes. The 
reduction was first made by Blume, in the original publication of 
the genus cited above, but he recognized two species, the first 
Cubilia blancoi, the second C. rumphii; it is very apparent that 
the two are identical, and that, so far as material at present 
available shows, the genus Cubilia is monotypic and is confined 
to the Philippines and the Moluccas. Blume had no Philippine 
material and interpreted Euphoria cubili Blanco solely from 
Blanco's description. Doctor Radlkofer, under date of July 5, 
1904, in discussing the identity of certain Philippine specimens 
with Cubilia blancoi and C. rumphii, writes as follows: 

Your Philippine plant seems to be quite identical with the Celebes one 
[coll. Koorders] and so Cubilia blancoi and C. rumphii will fall together, 
and the latter must become a synonym of the former. 

The differences indicated by Blume in the descriptions of what 



SAPINDACEAE 



339 



he took to be two species are certainly due only to translations 
of Blanco's descriptive terms in indicating the fruit and leaf 
characters. 

POMETIA Forster 

POMETIA PINNATA Forst. Char. Gen. (1776) 110. 
Irina glabra Blume Bijdr. (1825) 230. 

Dabanus (incl. Dawan batu, Dawan mera, Dawati puti) Rumph. Herb. 

Amb. 3: 31, 32, tt. 16, 17. 
Amboina, Negri lama, Robinson PL Rumph. Amb. 2, September 8, 1913, 
on river banks at an altitude of about 25 meters, locally known as tauwan 
itam; Kati-kati, Robinson PL Rumph. Amb. 3, October 7, 1913, in light 
forests, altitude about 90 meters. 

The reduction of Dabanus of Rumphius to Irina glabra Blume 
was made by Blume, Rumphia 3 (1847) 114; that is, Dawan batu 
Rumph. Herb. Amb. 3: 31, t. 17, to Irina glabra var. solida Blume 
and Dawan mera Rumph. Herb. Amb. 3: 32 to Irina glabra var. 
rubra Blume. Irina glabra Blume is a synonym of Pometia 
pinnata Forst. All three forms indicated by Rumphius are 
apparently referable to the widely distributed and variable 
Pometia pinnata Forst. Plate 1 6 represents a fruiting specimen 
and plate 17 a flowering specimen, the latter distinctly charac- 
teristic, the former rather crude, yet unmistakably a Pometia. 

JAG ERA Blume 

JAGERA SERRATA (Roxb.) Radlk. in Sitzb. Math.-Phys. Acad. Muench. 

8 (1878) 303. 

Sapindus serratus Roxb. Hort. Beng. (1814) 88 nomen nudum, L v l, 

Ind. ed. 2, 2 (1832) 284. 
Jagera speciosa Blume Rumphia 3 (1847) 155. 
Papaja silvestris minor Rumph. Herb. Amb. 1: 150, t. 53, f. 2. 

This species is not represented in our Amboina collections. 
Blume makes this reduction in the original description of Jagera 
speciosa, which is probably the correct disposition of Papaja 
silvestris minor Rumph. 

MISCHOCARPUS Blume 

M ISC HOC AR PUS FUSCESCENS Blume Rumphia 3 (1837-47) 169. 

Arbor palorum alba latifolia Rumph. Herb. Amb. 3: 99, t. 65, quoad 
/. A. 

This may or may not prove to be the correct disposition of 
the form that Rumphius described, the reduction following 
Miquel's suggestion. It is at least a representative of the 
Sapindaceae. 



340 



RUMPHIUS'S HERBARIUM AMBOINENSE 



DODONAEA Linnaeus 

DODONAEA VISCOSA (Linn.) Jacq. Enum. PI. Carib. (1760) 19. 

Ptelea viscosa Linn. Sp. PI. (1753) 118. 

Caryophyllaster litoreus Rumph. erb. Am b. : ! ' t. ->0. 
Amboina, Tandjong martafrons, Robinson PL Rumph. Amb. 1, October 
16, 1913, along the beach, locally known as chenki laut. The exact form, 
as determined by Doctor Radlkofer, is Dodonaea viscosa Jacq. var. vulgaris 
Benth., forma repanda Radlk. 

The reduction of Caryophyllus litoreus to Ptelea viscosa Linn, 
was first made by Linnaeus, in Stickman Herb. Amb. (1754) 17, 
Amoen. Acad. 4 (1759) 127, Syst. ed. 10 (1759) 898, Sp. PL 
ed. 2 (1762) 173, and transferred to Dodonaea as D. viscosa 
Linn., Mant. 2 (1771) 228, eleven years after Jacquin made the 
same transfer. Other names given by Hasskarl, Neue Schlussel 
(1866) 83, are Dodonaea burmanniana DC, D. dioica Roxb., D. 
triquetra Andr., and D. angustifolia Blanco, all of which are 
apparently synonyms of Dodonaea viscosa Jacq. 

HARPULLIA Roxburgh 

HARPULLIA ARBOREA (Blanco) Radlk. in Sitzb. Math.-Phys. Akad. 
Muench. 16 (1886) 404. 
Ptelea arborea Blanco Fl. Filip. (1837) 63. 
Metrosideros molucca fungosa Rumph. Herb. Amb. 3: 25. 
Amboina, Hitoe lama, Robinson PL Rumph. Amb. U, November 4, 1913, 
in forests at an altitude of about 50 meters, locally known as samar ayer. 
A form with larger leaflets, and, for the species, rather large fruits, is 
represented by Rel. Robins. 1601, from Hitoe messen, October 10, 1913, grow- 
ing on forested limestone hills at an altitude of about 150 meters. 

The identification has been made chiefly from the native name 
cited by Rumphius, that is, samar ayer, which also appears on 
one of the specimens cited above. The description given by 
Rumphius is entirely inadequate to warrant an identification of 
the form from it alone. It is to be noted, however, that the other 
two species described under the heading Metrosideros molucca 
by Rumphius have nothing to do with Harpullia, but one is 
Homalium foetidum Benth. and the other is indeterminable. 

BALSAMINACEAE 

I M PAT I ENS Linnaeus 
IMPATIENS BALSAM INA Linn. Sp. PL (1753) 938. 
Lacca herba Rumph. Herb. Amb. 5: 256, t. 90. 
Amboina, Binting, Robinson PL Rumph. Amb. 70, September 27, 1913. 
Lacca herba was reduced to Impatiens balsamina Linn, by 
Linnaeus, in Stickman Herb. Amb. (1754) 21, Amoen. Acad. 4 



RHAMNACEAE 



341 



(1759) 130, Syst. ed. 10 (1759) 1239, Sp. PL ed. 2 (1763) 1328, 
which is certainly the correct disposition of it, and which has been 
very generally followed by later authors. I consider that all of 
the forms indicated by Rumphius, / coccinea, II purpurea, III 
alba, and IV e Sina, are merely color forms of the widely dis- 
tributed and variable species. They have been referred by 
various authors to Impatiens coccinea Sims, Balsamina fasci- 
culata DC, B. tilo DC, and B. hortensis Desp.* 

RHAMNACEAE 

ZIZYPHUS Linnaeus 

ZIZYPHUS JUJUBA Lam. Encycl. 3 (1789) 318. 
Rhamnus jujuba Linn. Sp. PL (1753) 194. 
Malum indicum Rumph. Herb. Amb. 2: 117, t. 86. 
Amboina, Binting, Robinson PL Rumph. Amb. 267, November 18, 1913, 
from cultivated trees, locally known as vidara. 

This was reduced by Linnaeus to his Rhamnus jujuba, in Stick- 
man Herb. Amb. (1754) 9, Amoen. Acad. 4 (1759) 121, Sp. PL 
ed. 2 (1762) 282, in which he was followed by other authors until 
Lamarck transferred Rhamnus jujuba to Zizyphus, where it 
properly belongs. All authors who have cited Rumphius since 
Lamarck, refer it to Zizyphus jujuba (Linn.) Lam. Skeels, 
U. S. Dept. Agr. Bur. PL Ind. Bull. 208 (1911) 67, considers that 
Zizphus jujuba Lam. is invalidated by Z. jujuba Mill., Gard. 
Diet. (1768), and proposes to adopt for the plant commonly 
known as Zizyphus jujuba Lam. the name Z. mauritiana Lam. 
Judging from the data given by Rumphius, the three forms 
mentioned by him on page 118 as coming from Timor and Java 
are merely slight variants of this common and widely distributed 
species. 

COLU BR I N A f Richard 

COLUBRINA ASIATICA (Linn.) Brongn. in Ann. Sci. Nat. I 10 (1827) 
369. 

Ceanothus asiaticus Linti. Sp. PI. (1753) 196. >. 
Amara litorea Rumph. Herb. Amb. 5: 74, t. 39, f. 2. 
Amboina, Paso, Robinson PI. Rumph. Amb. 269, October 29, 1913, in 
thickets back of the beach, and at Ayer putre in similar habitat, August 
23, 1913. 

Amara litorea Rumph. has not previously been properly 
reduced. The figure is a good representation of this common and 
widely distributed Indo-Malayan strand plant. Suggested re- 

* See Hasskarl, Neue Schlussel (1866) 118. 

t Retained name, Vienna Code; Marcorella Neck. (1790) is older. 



342 RUMPHIUS'S HERBARIUM AMBOINENSE 



ductions by other authors have been Croton sp., by Burman f., 
ex Hasskarl, Neue Schliissel (1866) 96; some aurantiaceous 
plant, after Poiret, Hasskarl, 1. c. ; and Teysmann's suggestion to 
Hasskarl that it was Zizyphus timoriensis DC. 

VENTILAGO Gaertner 

VENTILAGO sp. 

Funis viminaiis Rumph. Herb. Amb. 5: 3, t. 2. 

Nothing that can be referred to the plant that Rumphius 
figures and describes occurs in our Amboina collections. The 
plant is manifestly a Ventilago, but its status must remain doubt- 
ful until more comprehensive collections are made in Amboina. 
Linnaeus, in Stickman Herb. Amb. (1754) 18, Amoen. Acad. 4 
(1759) 128, erroneously referred it to Securidaca volubilis Linn., 
with which, however, it has nothing in common. Willdenow, 
Sp. PL 1 (1797) 1106, reduced it to Ventilago maderaspatana 
Gaertn., in which he has been followed by all subsequent authors 
who have had occasion to cite the Rumphian figure. It is very 
improbable, however, that this Moluccan plant is identical with 
Ventilago maderaspatana Gaertn. Possibly it is the same as 
Ventilago cernua Tul., in Ann. Sci. Nat. Bot. IV 8 (1857) 123, 
which was described from specimens collected by Gaudichaud in 
Rawak Island, Moluccas. 

VITACEAE 

AMPELOCISSUS Planchon 

AMPELOCISSUS ARACH NOI DEA (Hassk.) Planch, in DC. Monog. Phan. 
5 (1887) 375. 

Cissus arachnoidea Hassk. Cat. Hort. Bogor. (1844) 166. 
Cissus blumeana Hassk. in Flora 25 (1842) Beibl. 39, non Span., 
nec Steud. 

Ampelopsis indica Blume Bijdr. (1825) 193, non Ampelocissus indica 
Planch. 

Labrusca molucca Rumph. Herb. Amb. 5: 452, t. 167. 

This rather characteristic species is not represented in our 
Amboina collections. The figure and the description, however, 
apply closely to Ampelocissus arachnoidea (Hassk.) Planch, and, 
for that matter, also to Ampelocissus martini Planch., which must 
be very closely allied to the former. Koorders, Exkurs. Fl. Java 
2 (1912) 557, seems first to have made this reduction of Labrusca 
molucca Rumph., which is manifestly the correct disposition of 
it. Linnaeus, in Stickman Herb. Amb. (1754) 24, Syst. ed. 
10 (1759) 942, erroneously referred it to Vitis indica Linn, and 
again, with even greater error, placed it under Vitis trifolia Linn., 
in Amoen. Acad. 4 (1759) 133. Loureiro, Fl. Cochinch. (1790) 



VITACEAE 



343 



155, referred it to Vitis labrusca Linn., but the Cochin-China 
form he described under the Linnean name was probably Am- 
pelocissits martini Planch, or A. arachnoidea Planch. 

CISSUS Linnaeus 

CISSUS QUADRANGULARIS Linn. Mant. 1 (1767) 39. 
Vitis quadrangularis Wall. Cat. (1832) no. 5992. 
Funis quadrangularis Rumph. Herb. Amb. 5: 83, t. Uh, f. 2. 

This characteristic species is not represented in our Amboina 
collections; it is, however, of local occurrence in many parts of 
Malaya, apparently here an introduced plant. Linnaeus cites 
the Rumphian name, Funis quadrangularis, Sp. PL ed. 2 (1763) 
1468, as a synonym of Menispermum crispum Linn., an error for 
Funis felleus Rumph., as the illustration 'indicated is t. UU, /. 1, 
which is a Tinospora. In the original description of Cissus qua- 
drangularis Linn., Funis quadrangularis Rumph. is cited as a 
synonym, this reduction certainly being the correct disposition 
of it. Most authors have quoted the Rumphian name and figure 
under Cissus quadrangularis Linn., a few under its synonym 
Vitis quadrangularis Wall. 

CiSSUS REPENS Lam. Encycl. 1 (1783) 31. 

Cissus cor data Roxb. Hort. Beng. (1814) 11, nomen nudum, Fl. Ind. 

ed. 2, 1 (1832) 407. 
Vitis repens W. & A. Prodr. (1834) 125. 

Funis crepitans I major Rumph. Herb. Amb. 5: 446, t. 164, f. 1. 
Amboina, Eri, Robinson PI. Rumph. Amb. 215, September 23, 1913, in 
thickets near the seashore, locally known as bunga tangong. 

Through confusion of Vitis alba Rumph. with t. 164>, /. 1, Lin- 
naeus originally reduced the above figure to Bryonia cordifolia 
Linn., in Stickman Herb. Amb. (1754) 24; it is manifest that 
he intended to cite 1. 166, both here and in Amoen. Acad. 4 (1759) 
133. In the second edition of the Species Plantarum (1762) 
170, he erroneously reduced Funis crepitans Rumph. to Vitis 
vitiginea Linn. Vahl, Symb. 3 (1794) 18, places it under Cissus 
latifolia Vahl, of which, however, it is not the type; Murray, 
Syst. (1774) 133, places it under Cissus sicyoides Linn., where 
it certainly does not belong; Willdenow, Sp. PL V (1797) 656, 
places it under Cissus latifolia Lam. ; and finally Roxburgh, FL 
Ind. ed. 2, 1 (1832) 407, places it under Cissus cordata Roxb., 
an exact synonym of Cissus repens Lam. Cissus repens Lam. 
was based on Neriam pidli Rheed., Hort. Malabar. 7: t. 48, and 
the form figured by Rheede appears to me to be specifically 
identical with the form figured by Rumphius as Funis crepitans 
I major and illustrated by the specimen cited above. 



344 



RUMPHIUS'S HERBARIUM AMBOINENSE 



CISSUS ARISTATA Blume Bijdr. (1824) 183. 

Oculus astaci Rumph. Herb. Amb. 5: 479, t. 178, f. 1. 
Amboina, Paso, Wae, and Koesoekoesoe sereh, Robinson PL Rumph. Amb. 
228, 229, 230, October and November, 1913, in thickets, sea level to an 
altitude of about 225 meters, locally known as siri barat. 

The identification of these specimens with Oculus astaci 
Rumph. is probably correct, although the stems, as presented by 
herbarium material, are not quadrangular, yet some of the 
branchlets are distinctly so. Hasskarl, Neue Schliissel (1866) 
149, referred it to Cissus glauca Roxb., which is supposed to be 
a synonym of Cissus repens Lam. Cissus aristata Blume has 
been reduced by Miquel and by Planchon to Cissus adnata Roxb., 
Fl. Ind. 1 (1820) 423, a species originally described from Indian 
material. The Amboina material seems to present some of the 
characters of Cissus assamica Craib, which like C. adnata Roxb. 
is widely distributed in the Indo-Malayan region, but its in- 
dumentum is of the adnata type. It is possible that the material 
should be referred to Cissus adnata Roxb. rather than to C. aris- 
tata Blume, but without access to the original material on which 
the various species were based, it is difficult to determine their 
exact relationships. The closely allied species involved are Cis- 
sus adnata Roxb., C. assamica Craib, C. aristata Blume, C. rotun- 
difolia Blume (C blumeana Steud.), and Cissus pyrrhodasys 
Miq., the last apparently being identical with Cissus assamica 
Craib var. pilosissima Gagnep., Not. Syst. 1 (1911) 353. 

VITIS QUA DR I CORN UTA Miq. Ann. Mus. Bot. Lugd. Bat. 1 (1863) 85. 
Funis crepitans II minor Rumph. Herb. Amb. 5: 446, t. 16b, f. 2. 

This form is not represented in our Amboina collections. 
It may be a variety of Cissus repens Lam., or it may be a closely 
allied form worthy of specific rank. Planchon, DC. Monog. 
Phan. 5 (1887) 506, repeats Miquel's description, under Cissus 
repens Lam., with the statement that Vitis diffusa Miq., V. mo- 
desta Miq., Vitis metziana Miq., and V. quadricornuta Miq. are 
either synonyms of Cissus repens Lam. or represent very closely 
allied species. The reduction of Funis crepitans II minor follows 
Miquel, but the plant, and for that matter Vitis quadricornuta 
Miq., is a true Cissus, although the transfer is not here definitely 
made in view of the uncertain status of the form described by 
Miquel. 

CISSUS sp.? 

Funis crepitans IV Rumph. Herb. Amb. 5: 447. 

Loureiro, Fl. Cochinch. (1790) 83, cites this as a synonym of 
Cissus trifoliata Lour., non Linn., a species of entirely uncertain 



VITACEAE 



345 



status. It must be a species of Columella or of Tetrastigma, if it 
belongs in the Vitaceae, as it probably does. I have not been able 
to refer to it any of our Amboina specimens, but further explora- 
tion of Amboina may yield material that will lead to a more 
definite determination of its status. 

COLUMELLA Loureiro 

COLUMELLA GEN ICU LATA (Blume) Merr. in Philip. Journ. Sci. 11 
(1916) Bot. 132. 
Cissus geniculata Blume Bijdr. (1825) 184. 
Cayratia geniculata Gagnep. in Not. Syst. 1 (1911) 345. 
Funis crepitans HI trifolia Rumph. Herb. Amb. 5: 447, t. 165. 
Amboina, Paso, Robinson PL Rumph. Amb. 226, October 29, 1913, on 
trees near the seashore. 

The specimen agrees perfectly with Rumphius's description 
and illustration and is manifestly the species commonly known as 
Cissus geniculata Blume, here called Columella geniculata Merr. 
Lamark, Encycl. 1 (1783) 31, reduced it to Cissus carnosa Lam.= 
C. trifolia K. Sch. of which, however, it is not the type. In this 
erroneous reduction he was followed by all authors who have had 
occassion to cite the Rumphian illustration ; namely, Vahl, Willde- 
now, Retzius, Poiret, Roemer and Schultes, Pritzel, and Hasskarl. 

COLUMELLA TRIFOLIA (Linn.) Merr. in Philip. Journ. Sci. 11 (1916) 
Bot. 134. 

Vitis trifolia Linn. Sp. PL (1753) 203. 
Cissus carnosa Lam. Encycl. 1 (1783) 31. 
Cissus crenata Vahl Symb. 3 (1794) 19. 
Cissus acida Murr. Syst. (1774) 133 (type!). 

Cissus trifolia K. Sch. in K. Sch. & Hollr. Fl. Kaiser Wilh. Land 
(1889) 71. 

Cayratia carnosa Gagnep. in Not. Syst. 1 (1911) 347. 

Folium causonis I album Rumph. Herb. Amb. 5: 450, t. 166, f. 2. 

Folium causonis II Rumph. Herb. Amb. 5: 450. 
Amboina, Batoe merah and Negri lama, Robinson PI. Rumph. Amb. 227, 
August and September, 1913, in thickets at low altitudes, locally known 
as kapiala. The same form is represented by Rel. Robins. 2U85, from 
Boeton, July 13, 1913. 

Folium causonis Rumph. was reduced by Linnaeus to Vitis 
trifolia Linn, in his Systema, ed. 10 (1759) 942, Sp. PL ed. 2 
(1762) 293, which, as Cissus trifolia K. Sch., or Columella tri- 
folia Merr., is the correct disposition of it. Of the various syn- 
onyms cited above it is the type and whole basis of but one, 
Cissus acida Murr., but is cited as a synonym in the original 
description of Cissus carnosa Lam. and of Cissus crenata Vahl. 
Roemer and Schultes, Syst. 3 (1818) 313, thought that it might 



346 



RUMPHIUS'S HERBARIUM AMBOINENSE 



represent Cissus trilobata Lam., which is a true Cissus and is 
known only from India. Miquel, Fl. Ind. Bat. 1 2 (1859) 602, 
placed it under Cissus cinerea Lam., which is apparently merely 
a pubescent form of Columella trifolia (Linn.) Merr., and 
in Ann. Mus. Bot. Lugd.-Bat. 1 (1863) 81 placed it, with 
doubt, under Vitis geniculata (Blume) Miq. = Columella genicu- 
lata Merr., a species that it certainly does not represent. The 
form described by Rumphius as Folium causonis II rubrum is 
apparently merely a color variant of Columella trifolia (Linn.) 
Merr., although Hasskarl, Neue Schlussel (1866) 145, thought 
that it might be Vitis geniculata Miq. 

Columella Lour., Fl. Cochinch. (1790) 85, has been accepted 
by me as the proper generic name for this group, although the 
recognition of it invalidates Columellia Ruiz & Pavon (1794) 
of the Columelliaceae. I am of the opinion that Gagnepain * 
was justified in segregating from Cissus as a distinct genus, 
those forms placed by Planchon in the subgenus Cayratia, dif- 
fering from him only in the selection of the generic name. 

TETRASTIGMA Planchon 

TETRASTIGM A sp. 

Folium causonis ill litoreum Rumph. Herb. Amb. 5: 450. 

No Tetrastigma occurs in our Amboina collections, yet from 
the description given by Rumphius, I have little doubt that 
the form described as Folium causonis III litoreum is referable 
to this genus. Hasskarl, Neue Schlussel (1866) 145, thought 
that it might be referable to Vitis geniculata Miq. var. grosse- 
serrata Miq. 

LEEA Royen 

LEEA AEQUATA Linn. Mant. 1 (1767) 124. 

Frutex aquosus femina Rumph. Herb. Amb. 4: 103, t. J+5. 
Amboina, Batoe merah, Robinson PL Rumph. Amb. 565, July 20, 1913, 
in rocky soil at low altitudes. 

Burman f., Fl. Ind. (1768) 78, erroneously placed this under 
Aralia chinensis Linn. Willdenow, Sp. PI. 1 2 (1797) 1177, re- 
ferred it to Leea sambucina Willd., in which he was followed by 
numerous authors, Persoon, Roxburgh, Roemer and Schultes, Don, 
Dietrich, and others. Miquel, Ann. Mus. Bot. Lugd. Bat. 1 
(1863) 98, reduced it to Leea aequata Linn., which is certainly 
the correct disposition of Frutex aquosus femina Rumph. The 
figure is poor, and from it alone it would be difficult to determine 



* Gagnepain, F. Un genre meconnu : classification des Cissus et Cayratia. 
Not. Syst. 1 (1911) 339-362. 



VITACEAE 



347 



which of several species of Leea was intended ; but the description 
is unmistakably that of Leea aequata Linn., not of Leea sambu- 
cina Willd., especially in the description of the leaflets as "su- 
perne pilis hirta, inferne arenulosa," "arenulosa" manifestly 
referring to the numerous glands on the lower surface that are 
so very characteristic of Leea aequata Linn. 

LEEA ACULEATA Blume Bijdr. (1825) 197; Miq. Ann. Mus. Bot. Lugd.- 
Bat. 1 (1863) 99. 
Leea spinosa Spreng. Syst. 1 (1825) 670 p. p., quoad cit. "Ins. 
Molucc." 

Leea aculeata Blume var. moluccana Miq. 1. c. 
Leea serrulata Miq. 1. c. 

Frutex aquosus mas Rumph. Herb. A . b. 4: 102, t. Uh> 
Amboina, Mahija and town of Amboina, Robinson PL Rumph. Amb. 231, 
232, July 23 and August 7, 1913, along river banks and edges of clearings, 
altitude 7 to 200 meters, locally known as kayu baduri. 

The specimens certainly represent Frutex aquosus mas Rumph. ; 
they are also undoubtedly referable to the species described by 
Blume as Leea aculeata Bl. and later described by Blanco from 
Philippine material as a distinct species under the same specific 
name. The trunk is supplied with short spines, but ordinary 
herbarium material rarely presents these as the branchlets are 
nearly always unarmed. Chiefly on account of the spiny stems 
as depicted by Rumphius, Linnaeus referred Frutex aquosus mas, 
with doubt, to Aralia chinensis Linn., in Stickman Herb. Amb. 
(1754) 16, Amoen. Acad. 4 (1759) 127, Syst. ed. 10 (1759) 967, 
Sp. PI. ed. 2 (1762) 393, in which he was followed by Lamarck, 
Loureiro, Willdenow, Burman f., and other authors. Koste- 
letzky, Allg. Med.-Pharm. Fl. 5 (1836) 1981, referred it to Aralia 
spinosa Linn. De Candolle, Prodr. 4 (1830) 259, expressed the 
opinion that the Rumphian figure represented a species of Leea, 
rather than an Aralia; and finally Miquel, Ann. Mus. Bot. Lugd. 
Bat. 1 (1863) 99, reduced Frutex aquosus mas Rumph. to Leea 
aculeata Blume, which is apparently the correct disposition of it. 
C. B. Clarke, Journ. Bot. 10 (1881) 105, expresses the opinion 
that Frutex aquosus mas Rumph. Herb. Amb. 4: 102, t. UU, does 
not represent Leea aculeata Blume; but it is apparent that he 
misinterpreted Blume's species, because he puts it in the section 
with red flowers, while in Leea aculeata Blume the flowers are 
greenish-white or white. The name Leea spinosa Spreng. is 
properly a synonym of Aralia chinensis Linn. ; Sprengel ap- 
parently intended to refer here only the Amboina reference from 
Rumphius, but does not so state. Following the short descrip- 
tion he merely cites "Ins. Molucc. China (Aralia chinensis L.)." 



348 



RUMPHIUS'S HERBARIUM AMBOINENSE 



ELAEOCARPACEAE 

ELAEOCARPUS Linnaeus 

Five or six species of Elaeocarpus are described by Rumphius 
under various names, such as Blimbingum sylvestre Rumph., 
Ganitrus, Ganitrum oblongum, Lignum momentaneum, and 
Arbor rediviva; while Hasskarl, Neue Schiissel (1866) 49, 50, 
suggests that Sicchius femina Rumph. Herb. Amb. 3: 41, t. 22, 
and Carbonaria femina Rumph., 1. c. 53, may be species of Elaeo- 
carpus. The last is certainly not referable to the genus. 

Five species of this genus have been described from Amboina. 
These are Elaeocarpus oppositif olius (DC.) Miq. and E. frutico- 
sus Roxb. (from the Moluccas, probably from Amboina), both 
very imperfectly described, indicated as having opposite leaves, 
and certainly the same as E. edulis T. & B. ; E moluccanus 
SchefT. ; and E. treubii Hochr. The last two were described from 
specimens cultivated in the botanic garden at Buitenzorg, Java, 
and E. treubii Hochr. may not have originated in Amboina. 
The Robinson collection presents three species, one of which is 
certainly the same as Elaeocarpus edulis T. & B. and with equal 
certainty is the same as Blimbingum sylvestre Rumph., but the 
other two I cannot definitely refer to any described species. 
While both have doubtfully been referred to forms figured and 
described by Rumphius, the identity of neither with the 
Rumphian plants can be considered certain. 

ELAEOCARPUS OP POSIT I FOLI US (DC.) Miq. Fl. Ind. Bat. 1 2 (1858) 211. 

Aceratium oppositif olium DC. Prodr. 1 (1824) 519. 
Elaeocarpus fruticosus Roxb. Fl. Ind. ed. 2, 2 (1832) 600. 
Elaeocarpus excavatus Reinw. ex Koord. in Lorenz Nova Guinea 8 
(1907) 174. 

Biimbingum sylvestre Rumph. Herb. Amb. 4: 138, t. 73. 
Amboina, Negri lama, Robinson PL Rumph. Amb. 372, September 8, 1913; 
Way tommo, Robinson PL Rumph. Amb. 373, August 17, 1913, PL Rumph. 
A?nb. 190, July 23, 1913 (detached mature fruits only), river banks, 
altitude 15 to 40 meters. 

Loureiro, Fl. Cochinch. (1790) 69, was entirely wrong in 
referring Blimbingum sylvestre of Rumphius to his new genus 
and species, Cylindria rubra. The genus is of entirely uncertain 
status, has nothing to do with Elaeocarpus, and must be inter- 
preted from Cochin-China material. Elaecarpus edulis Teysm. 
& Binn. was based on Amboina specimens, but these authors do 
not indicate the identity of Blimbingum sylvestre with their 
species. The actual specimens, cited above, agree with Rumph- 
ius's dscription and rather poor figure and bear the native name 



ELAEOCARPACEAE 



349 



tagorela; the Amboinese name cited by Rumphius is tagorela 
abbal. The mature fruits are red, about 4 cm long, 2 cm wide, 
3-angled, apex acute or acuminate, the pericarp rather peculiarly 
produced at the base, giving the fruit the truncate appearance 
represented in Rumphius's figure. Teysmann gave the native 
names as bliembieng-oetan, tagorela, and kakarja. The type 
of Aceratium oppositi folium DC. was from Amboina, and the 
description conforms with the characters of the species commonly 
known as Elaeo carpus edulis T. & B., the type of which was also 
from Amboina. The type of Elaeocarpus oppositifolius Roxb. 
was from the Moluccas, very likely from Amboina, and Rox- 
burgh's short description also conforms with Elaeocarpus edulis 
T. & B. Elaeocarpus excavatus Reinw., a herbarium name 
published by Koorders, was also based on an Amboina speci- 
men. The species is in cultivation in the botanic garden, Bui- 
tenzorg, Java, and occurs also in New Guinea. 

ELAEOCARPUS RUM PHI I sp. nov. § Monocera. 

Arbor red i viva Rumph. Herb. Amb. 3: 165, t. lOJt? 
Amboina, between Soja and Hatalai and at Hitoe messen, Robinson PL 
Rumph. Amb. 370 (type), October 24 and November 1, 1913, in forests, al- 
titude 175 to 350 meters. 

Arbor circiter 12 m alta, inflorescentiis exceptis glabra ; foliis 
oblongis, coriaceis, longe petiolatis, usque ad 20 cm longis, 
obscure obtuse acuminatis, basi rotundatis vel leviter cordatis, 
margine obscure crenatis, nervis utrinque 8 ad 10, prominen- 
tibus; racemis numerosis, pubescentibus, floribus longe pedicel- 
latis, sepalis lanceleolatis, pubescentibus, circiter 10 mm longis, 
petalis sepalis aequantibus, intus densissime retrose hirsutis, 
apice subacutis leviter parce lobatis haud fimbriatis. 

A tree about 12 m high, quite glabrous except the inflorescence. 
Branches reddish-brown, terete, smooth, the ultimate ones 6 
to 8 mm in diameter. Leaves alternate, coriaceous, green and 
shining when dry, 15 to 20 cm long, 5 to 7.5 cm wide, gradually 
narrowed upward to the obscurely blunt-acuminate apex, the 
base rather broad, somewhat abruptly rounded or slightly cor- 
date, margins distantly and rather obscurely crenate; lateral 
nerves 8 to 10 on each side of the midrib, prominent, anasto- 
mosing, the primary reticulations slender, subparallel, the ulti- 
mate ones rather close, distinct, the whole lower surface with 
scattered, minute dark-colored, roundish glands or gland-like 
bodies; petioles reddish-brown, 6 to 7 cm long. Racemes axil- 
lary, about 15 cm long, uniformly pubescent with short, grayish 
hairs as are the pedicels and sepals. Pedicels about 1.5 cm long. 



350 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Sepals 5, lanceolate, about 10 mm long, 3 mm wide, narrowed 
upward, subacute. Petals as long as the sepals, oblong-lanceo- 
late, outside densely pubescent with pale-brownish, shining, ap- 
pressed hairs, inside uniformly and densely hirsute with reflexed 
hairs, the apical 2 mm cut into few narrow lobes, usually two 
lateral slender ones on each side with a central somewhat stouter 
one. Stamens about 45, the filaments scabrid, 2 to 2.5 mm long ; 
anthers linear, scabrid, 4.5 to 5 mm long including the slender 1 
to 1.5 mm long awn that terminates one cell. Ovary ovoid, 
densely pubescent, 2-celled; style 4 to 5 mm long, pubescent 
below. 

It is by no means certain that this is Arbor rediviva of Rumph- 
ius, although the plant figured and described by Rumphius is 
manifestly an Elaeocarpus. In the species above described the 
leaves are larger than in Arbor rediviva, rounded or cordate at 
the base, not acute, and do not present the peculiar protuberances 
(galls?) shown by Rumphius. It was by Loureiro, FL Cochinch. 
(1790) 663, referred to Dicalyx cochinchinensis Lour., but this is 
a species of Symplocos and judging from the description is 
entirely different from Arbor rediviva Rumph. 

ELAEOCARPUS AM BOf N ENSIS sp. nov. § Ganitrus. 
Ganitrus Rumph. Herb. Amb. 3: 160, t. 101? 
Amboina, Paso, near the coast, Robinson PL Rumph. Amb. 371 (type), 
November, 25, 1913. 

Arbor circiter 12 m alta partibus junioribus inflorescentiisque 
parce pubescentibus exceptis glabra; foliis alternis, anguste 
oblongis, chartaceis, nitidis, usque ad 15 cm longis, utrinque 
subaequaliter angustatis, basi acutis, apice obtusis, margine cre- 
nulatis, nervis utrinque circiter 13, prominentibus ; racemis e 
ramis defoliatis, tenuibus, 10 ad 12 cm longis ; floribus numerosis, 
tenuiter pedicellatis, sepalis anguste lanceolatis, acuminatis, par- 
cissime pubescentibus, circiter 11 mm longis, petalis aequilongis, 
oblongo-lanceolatis, apice usque ad § fissis, basi ad margine dense 
puberulis exceptis glabris ; ovario 5-loculare. 

A tree about 12 m high, glabrous except the slightly appressed- 
pubescent younger parts and inflorescence. Branches terete, 
reddish-brown, glabrous, the younger ones somewhat angled, 
brownish-olivaceous, minutely and rather sparingly pubescent. 
Leaves alternate, narrowly oblong, firmly chartaceous, suboliva- 
ceous, of about the same color on both surfaces and shining when 
dry, 12 to 15 cm long, 3 to 5 cm wide, subequally narrowed to the 
acute base and to the blunt apex, the tip sometimes slightly refuse 
and minutely apiculate, margins distinctly crenulate; lateral 



ELAEOCARPACEAE 



351 



nerves about 13 on each side of the midrib, slender but prominent, 
curved, anastomosing, the reticulations distinct ; petioles slender, 
about 1 cm long, the younger ones slightly pubescent. Racemes 
numerous, solitary, spreading, from the branches below the leaves 
in the axils of fallen leaves, 10 to 12 cm long, slender, the rachis 
and pedicels more or less pubescent with short, appressed, pale- 
gray hairs. Flowers 20 to 25 or more in each raceme, the buds 
lanceolate, acuminate, whitish when fresh, brown when dry, the 
opened flowers greenish, the pedicels slender, about 1.5 cm long. 
Sepal 5, narrowly lanceolate, acuminate, externally slightly pub- 
escent, about 11 mm long, 2 to 2.5 mm wide. Petals as long as 
the sepals, oblong-lanceolate, the upper two-fifths cut into from 
15 to 20 slender fimbriae, these united into 5 or 6 primary divi- 
sions, quite glabrous except the densely pubescent or puberulent 
margins in the lower part. Stamens about 50, the filaments 
slender, scabrid, 1 to 1.5 mm long; anthers linear, 5 to 6 mm long 
including the slender, solitary, bristle-like, 1 mm long awn that 
tips one of the cells. Ovary ovoid, densely pubescent, somewhat 
sulcate, 5-celled; style 5 to 6 mm long, pubescent below. 

This species is very closely allied to the Philippine Elaeocarpus 
dolichopetalus Merr., from which it is distinguished by its blunt, 
not acuminate leaves, its longer racemes, longer pedicels, and 
smaller flowers. 

It is by no means certain that it is the same as Ganitrus of 
Rumphius. The figure of Ganitrus presents relatively shorter, 
fewer-nerved, rather differently shaped leaves, and relatively 
shorter, fewer-flowered racemes. It was apparently drawn from 
Amboina specimens, although Rumphius includes in the descrip- 
tion specimens from other parts of the Malay Archipelago. 

Historically, Ganitrus was first reduced by Linnaeus to his 
Elaeocarpus serratus, in Stickman Herb. Amb. (1754) 13, Amoen. 
Acad. 4 (1759) 124, Syst. ed. 10 (1759) 1075, Sp. PI. ed. 2 (1762) 
734, in which he was followed by Burman f ., Loureiro, Willdenow, 
and Lamarck. However, Elaeocarpus serratus Linn., Sp. PL 
(1753) 515, was based on Indian and Ceylon plants, and has 
nothing to do with the plant described by Rumphius. Gaertner, 
Fruct. 2 (1791) 271, t. 139, f. 6, takes his generic name Ganitrus 
from Rumphius, and refers to Ganitrus sphaerica Gaertn. the 
figure and description of Rumphius mentioned above. Gaertner's 
description was from an actual specimen, and his species must be 
interpreted from the material described. Gaertner's species is 
probably the same as the one later described by Roxburgh as 
Elaeocarpus ganitrus Roxb., Fl. Ind. ed. 2, 2 (1832) 592, who 
refers to his species the Rumphian plant and takes his specific 



352 



RUMPHIUS'S HERBARIUM AMBOINENSE 



name from Rumphius. However, the species actually described 
by Roxburgh was based on Indian material and manifestly is 
not the Rumphian plant. The proper name for this Indian 
species is apparently Elaeocarpus sphaericus (Gaertn.) K. Sch. 
(E. ganitrus Roxb., non Ganitrus Rumph.). The species de- 
scribed above should be compared critically with Elaeocarpus 
angustifolius Blume (Aceratium ganitri Hassk.), to which it 
is manifestly allied. 

ELAEOCARPUS OBLONGUS Gaertn. Fruct. 1 (1791) 202, t. US, f. 3. 
Ganitrum oblongum Rumph. Herb. Amb. 3: 163, t. 102? 

This species is not represented in our Amboina collections. 
Rumphius states that the form he described was found in Am- 
boina, Celebes, and Bali. Gaertner placed it under Elaeocarpus 
oblongus in the original description of that species, but I have no 
means of determining whether or not the specimens he had 
before him were identical with the form Rumphius described and 
figured. He does not state the origin of his material. Lamarck, 
Encycl. 2 (1788) 604, referred it tentatively to Elaeocarpus integ- 
rifolius Lam., but his type was from the Isle of France and is 
certainly not the same as the Moluccan form. Willdenow, Sp. 
PL 2 2 (1799) 1170, followed Lamarck in his disposition of it. 
Hasskarl, Neue Schlussel (1866) 62, suggests that it is Elaeocar- 
pus macrophyllus Blume, which is very improbable. Pending a 
critical revision of the genus or at least of the Indo-Malayan 
species, it seems best to leave it under Elaeocarpus oblongus 
Gaertn. I have seen no authentic species of Gaertner's species, 
do not know of what country it is a native, and strongly suspect 
that current interpretations of it are merely approximate; per- 
haps the modern conception of the species is based more on 
Rumphius's figure than on Gaertner's actual specimens. 

ELAEOCARPUS sp.? 

Lignum momentaneum Rumph. Herb. Amb. 3: 164, t. 10S. 

This reduction is suggested by Hasskarl, Neue Schlussel 
(1866) 62, following Savigny's note in Lamarck, Encycl. 4 (1798) 
693, under the Rumphian name pagamat. There is little in the 
description or in the figure to indicate an Elaeocarpus, and if 
the infructescence is drawn correctly, it certainly is not a rep- 
resentative of this genus, although Rumphius compares the 
fruits with his Ganitrus, which is an Elaeocarpus. Its status 
should be determinable from continued field work in Amboina, 
as Rumphius states that it was common in the Moluccas and cites 
the native names pagamatta and pegang matta for Amboina and 
sal for Ternate. 



GONYSTYLACEAE— TILIACEAE 



353 



GONYSTYLACEAE 

GONYSTYLUS Teysmann and Binnendyck 

GONYSTYLUS BANCANUS (Miq.) Baill. ex Hook. f. & Jackson Index 
Kewensis 2 (1895) 1055. 
Aquilaria bancana Miq. PI. Ind. Bat. Suppl. (1860) 141, 355. 
Gony stylus miquelianus T. & B. in Bot. Zeit. 20 (1862) 265. 
Agallochum spurium Rumph. Herb. Amb. 2: 40. 

Teysmann and Binnendyck reduced Agallochum spurium 
Rumph. to Gony stylus miquelianus T. & B., in Miq. Ann. Mus. 
Bot. Lugd. Bat. 1 (1864) 133, where the species is minutely de- 
scribed and figured. It is impossible definitely to determine from 
Rumphius's description whether the plant he had in mind was 
G. bancanus Baill. or was one of the other species of the genus. 
The fact that a portion of his material came from Borneo leads 
one to suspect that he may have had one of the Bornean species, 
G. affinis Radlk., G. borneensis Gilg, G. pluricornis Radlk., or 
G. calophyllus Gilg. 

Agallochum spurium album Rumph., very briefly described in this 
chapter, is indeterminable from any data at present available. 
Agallochum spurium III, merely mentioned, is Excoeearia agallocha 
Linn., later described and figured by Rumphius, Herb. Amb. 
2: 237, t. 79, 80 (see p. 327). 

TILIACEAE 

CORCHORUS Linnaeus 

CORCHORUS CAPSULARIS Linn. Sp. PI. (1753) 529. 

Ganga sativa Rumph. Herb. Amb. 5: 212, t. 78, f. 1. 

This spcies is not represented in our Amboina collections, but 
Rumphius's figure is unmistakably Corchorus capsularis Linn. 
It was first reduced by Linnaeus, in Stickman Herb. Amb. (1754) 
21, Amoen. Acad. 4 (1759) 130, Syst. ed. 10 (1759) 1079, all 
later authors concurring in this reduction. 

CORCHORUS OLITORIUS Linn. Sp. PL (1753) 529. 

Ganga agrestis Rumph. Herb. Amb. 5: 213 (haud t. 78, f. 2). 

This species is not represented in the Amboina collections. 
The description is unmistakably that of Corchorus olitorius Linn. 
The figure, however, is certainly no Corchorus, but I am unable 
to suggest what species was intended; it does not agree at all 
with the description in either its leaf or its fruit characters. 
The figure was referred by Linnaeus to Corchorus olitorius in 
Stickman, Herb. Amb. (1754) 21, Amoen. Acad. 4 (1759) 130. 

144971 23 



♦ 



354 



RUMPHIUS'S HERBARIUM AMBOINENSE 



COLUMBIA Per soon 

COLUMBIA SUBOBOVATA Hochr. PI. Bogor. Exsicc. (1904) 25. 

Restiaria nigra Rumph. Herb. Amb. 3: 188. 

Perticaria tertia Rumph. Herb. Amb. 3: 189, t. 120. 

Perticaria tertia latifolia Rumph. Herb. Amb. 3: 189. 
Amboina, Negri lama, Robinson PI. Rumph. Amb. 279, September 8, 1913, 
in light woods, altitude 10 meters, locally known as hunot; Amboina (town), 
Robinson PI. Rumph. Amb. 278, July 30, 1913, river sides at 5 meters 
altitude, locally known as morong itam; Paso, Robinson PI. Rumph. Amb. 
280, near the seashore, October 29, 1913, locally known as morong mera and 
hunut puti. 

There is very little doubt that the three Rumphian descrip- 
tions, cited above, represent the same species, and that all 
are Columbia (Diplophr actum) subobovata Hochr. The species 
presents great variation in its vegetative characters, the leaves 
varying from 8 to 35 cm in length and from 5 to 18 cm in width, 
all intermediate sizes being found on the same specimen. Like- 
wise the leaf bases on the same specimen vary from perfectly equi- 
lateral to strongly inequilateral. The figure given by Rumph- 
ius for Perticaria tertia is very poor, and one would never 
suspect that it was intended for a species of Columbia. The 
description, however, is unmistakable. Loureiro, Fl. Cochinch. 
(1790) 639, referred, with doubt, Restiaria nigra Rumph., Herb. 
Amb. 3: 188, to Restiaria cor data Lour. Loureiro's species, how- 
ever, is supposed to be an Uncaria, is certainly not the same as 
the Rumphian one, and must be interpreted from Cochin-China 
material. The proper disposition of Perticaria has not before 
been indicated.* 

TRIUMFETTA Linnaeus 

TRIUMFETTA BARTRAM I A Linn. Syst. ed. 10 (1759) 1044. 

Triumfetta rhomboidea Jacq. Enum. PI. Carib. (1762) 22, Stirp. Am. 

Hist. (1763) 147, t. 90. 
Lappago amboinica sylvestris Rumph. Herb. Amb. 6: 60 (haud t. 

25, f. 2). 

Amboina, town of Amboina and Gelala, Robinson PL Rumph. Amb. 277, 
August, 1913, along the beach and in waste places at low altitudes. 

Lappago amboinica Rumph. Herb. Amb. 6: 59, t. 25, f. 2, is 
certainly Urena lobata Linn. The figure is poor, but the drawing 
of the fruits and the description are unmistakably Urena, while 
the drawing of the flowers on the same plant are likewise 

* For a further discussion of the status of Restiaria nigra Rumph. 
see Trichospermum quadrivalve Merr. in Philip. Journ. Sci. 11 (1916) 
Bot. 289. 



MALVACEAE 



355 



unmistakably those of Triumfetta. The figure has by some 
authors been referred to Urena, by others to Triumfetta. The 
description of Lappago amboinica sylvestris is certainly that of 
a Triumfetta, Hasskarl, "Neue Schliissel (1866) 163, has re- 
ferred it to Triumfetta rotundifolia Lam., but Lamarck's species 
is Indian not Malayan. 

The oldest name for this species is Bartramia indica Linn. Sp. 
PI. (1753) 389, but the specific name is invalidated in Triumfetta 
by Triumfetta indica Lam. Triumfetta bartramia Linn, was 
based on Bartramia indica Linn. Sp. PI. (1753) 389, with the 
addition of a reference to "Rumph. Amb. 3: t. 119" which is 
Commersonia bartramia (Linn.) Merr. I interpret the type of 
Bartramia indica Linn, as Fl. Zeyl. 17 U, which according to 
Thwaites, Fl. Ceyl. 1 (1893) 180, is Triumfetta rhomboidea Jacq., 
who also states that : 

The name T. Bartramia has priority (1762), but Linnaeus may have 
included in it more than T. rhomboidea Jacq. as now understood. 

MALVACEAE ' 

ABUTILON Tournefort 

ABUTILON INDICUM (Linn.) Sweet Hort. Brit. (1827) 54. 

Sida indica Linn. Cent. PI. 2 (1756) 26, Amoen. Acad. 4 (1759) 324. 
Abutilon laeve Rumph. Herb. Amb. 4: 31, t. 11. 

This species is not represented in our Amboina collections, but 
Rel. Robins. 251 8 from Boeleleng, Bali, is Abutilon indicum 
Sweet. Abutilon laeve Rumph. was originally considered by 
Linnaeus to represent a variety of Sida abutilon Linn., in Stick- 
man Herb. Amb. (1754) 15, Amoen. Acad. 4 (1759) 125. It 
was first reduced to Sida indica Linn, by Loureiro, Fl. Cochinch. 
(1790) 414, which as Abutilon indicum (Linn.) Sweet is cer- 
tainly the correct disposition of it. By some authors it is 
considered to represent Sida indica Sweet var. populifolia (Lam.) 
Mast, in Hook. f. Fl. Brit. Ind. 1 (1874) 326, while Roxburgh, 
Fl. Ind. ed. 2, 3 (1832) 178, erroneously placed it under Sida 
abutilon Linn.= Abutilon avicennae Gaertn. 

ABUTILON HIRTUM (Lam.) Sweet Hort. Brit. (1827) 53. 

Sida hirta Lam. Encycl. 1 (1783) 7. 

Sida pilosa L'Herit. Stirp. (1784-85) 130. 

Abutilon hirsutum Rumph. Herb. Amb. 4: 29, t. 10. 

This species is not represented in our Amboina collections. 
The Rumphian description and plate were cited by Lamarck in 



356 



RUMPHIUS'S HERBARIUM AMBOINENSE 



the original description of Sida hirta, but the actual type of the 
species was a plant collected by Sonnerat. Linnaeus originally 
considered the plate to represent a variety of Sida abutilon Linn., 
in Stickman Herb. Amb. (1754) 15, Amoen. Acad. 4 (1759) 125; 
while Burman f., Henschel, and Murray erroneously placed it 
under Sida asiatica Linn. By other authors it has been placed 
under Abutilon graveolens Wight & Arn. (Sida graveolens 
Roxb.) , to which Abutilon hirtum has been reduced as a variety,* 
but even if varietal forms be recognized, the specific name, 
hirtum, being much older, should be retained. 

Abutilon litoreum Rumph. Herb. Amb. 4: 33, briefly described 
in the same chapter with Abutilon laeve= Abutilon indicum 
Sweet, was thought by Hasskarl, Neue Schlussel (1866) 73, 
possibly to represent Abutilon albescens Miq. This suggested 
reduction of Abutilon litoreum is possibly correct, but the definite 
identification of the form described by Rumphius must await 
further botanical exploration of the Moluccas. The plant is 
undoubtedly an Abutilon and may be a form of A. indicum Sweet. 

SIDA Linnaeus 

SIDA ACUTA Burm. f. Fl. Ind. (1768) 147. 

Sida carpinifolia Linn. f. Suppl. (1781) 307. 
Sida scoparia Lour. Fl. Cochinch. (1790) 414. 

Sigalurium II longifolium Rumph. Herb. Amb. 6: 45, t. 18, f. 2. 

Amboina, Lateri, Robinson PI. Rumph. Amb. 500, August 25, 1913, altitude 
about 100 meters. 

The figure given by Rumphius is unmistakably Sida acuta 
Burm. f., and Burman f. cites it in the original publication of 
his species, although his description is based primarily on 
actual specimens. Linnaeus erroneously referred it to Sida 
spinosa Linn., in Stickman, Herb. Amb. (1754) 26, Amoen. 
Acad. 4 (1759) 134, Syst. ed. 10 (1759) 1145. Loureiro, Fl. 
Cochinch. (1790) 414, cites it as representing his Sida scoparia. 
Most recent authors have followed Burman f ., which is certainly 
the correct disposition of Sigalurium longifolium. 

SIDA RETUSA Linn. Sp. PL ed. 2 (1763) 961. * 

Sigalurium I rotundum s. vulgare Rumph. Herb. Amb. 6: 44, t. 19. 

This species is not represented in our Amboina collections. 
The figure is unmistakably Sida retusa Linn., and it is cited in 
the original description of that species, although it is not the 
type. It was originally reduced by Linnaeus to Sida alnifolia 
Linn., in Stickman Herb. Amb. (1754) 26, Amoen. Acad. 4 



* Masters in Hook. f. Fl. Brit. Ind. 1 (1874) 327. 



MALVACEAE 



357 



(1759) 134, Syst. ed. 10 (1759) 1145, where it manifestly does 
not belong. 

SIDA CORDIFOLIA Linn. Sp. PL (1753) 684. 

Abutilon montanum Rumph. Herb. Amb. 4: 32. 

This species is not represented in our Amboina collections, 
but Rel. Robins. 2512, from Boeleleng, Bali, is a form of this 
common and widely distributed species. The reduction of Abu- 
tilon montanum to Sida cordifolia follows Hasskarl, Neue Schlus- 
sel (1866) 73, and is almost certainly the correct disposition 
of it. 

The form from the Cape of Good Hope, indicated by Rumphius, 
1. c, as Abutilon montanum e Capite bonae spei, suggested by Hasskarl, 
Neue Schlussel (1866) 73, as possibly Sida triloba Cav. or S. 
sonneratiana Cav. = Abutilon sonneratianum Sweet, may be iden- 
tifiable from a study of the South African Malvaceae in Sida, 
Abutilon, and other allied genera. It may be one of the species 
that Hasskarl has indicated, but again it may be quite different 
from both of these. 

SIDA spp.? 

Sigalurium rotundum silvestre Rumph. Herb. Amb. 6: 45. 
Sigalurium III album Rumph. Herb. Amb. 6: 45. 

The descriptions are too indefinite to warrant accurate de- 
termination of these two forms. Hasskarl, Neue Schlussel 
(1866) 160, suggests that the former may be Sida carpinoides 
~DC.=Malvastrum coromandelianum (Linn.) Garcke (M. tricus- 
pidatum G. Gray) and that the latter may be Sida alba Linn., 
but both of these suggested determinations are probably wrong. 

URENA Linnaeus 

URENA LOBATA Linn. Sp. PL (1753) 692. 

Lappago amboinica Rumph. Herb. Amb. 6: 59, t. 25, f. 2. 
Amboina, Caju poeti, Robinson PL Rumph. Amb. 4.95, August 2, 1913, 
roadsides, etc., up to an altitude of 350 meters. 

So far as the Rumphian figure and description go, Urena lobata 
includes the forms described as I laciniata and II latifolia, while 
III silvestris and the drawings of the attached flowers are Trium- 
fetta bartramia Linn, (see p. 354). The descriptions quoted 
above and the figure, excepting the flowers, are unmistakably 
Urena lobata Linn. On account of the mixture of two entirely 
different species in the drawing, the plate has by some been cited 
under Triumfetta, by others under Urena. The plate and de- 
scription, for the most part, are Urena lobata Linn., although 
originally reduced by Linnaeus, in Stickman Herb. Amb. (1754) 



358 



RUMPHIUS'S HERBARIUM AMBOINENSE 



26, Amoen. Acad. 4 (1759) 134, to Urena sinuata Linn., but cor- 
rected in his Systema, ed. 10 (1759) 1148, to Urena lobata. 
Other names involved in the reduction are Urena lappago Sm. 
and U. heterophylla Lam. 

ABELMOSCHUS Medikus 

ABELMOSCH US MOSCHATUS Medik. Malv. (1787) 46. 
Hibiscus abelmoschus Linn. Sp. PL (1753) 696. 
Granum moschatum Rumph. Herb. Amb. 4: 38, t. 15. 
Amboina, Toelehoe, Robinson PL Rumph. Amb. 492, November 25, 1913, 
in grasslands, altitude about 15 meters, locally known as daun kasturi. 

The original reduction of Granum moschatum to Hibiscus 
abelmoschus Linn, was made by Linnaeus, in Stickman Herb. 
Amb. (1754) 15, Amoen. Acad. 4 (1759) 126, Syst. ed. 10 (1759) 
1149, Sp. PL ed. 2 (1763) 980, which is certainly the correct 
disposition of it, although now it is very generally recognized as 
representing a distinct genus and is classified as Abelmoschus 
moschatus Medik. The Amboina specimen cited above is much 
more pubescent than the commoner forms of the species, but 
probably belongs here. 

ABELMOSCHUS MINDANAENSIS Warb. in Perk. Frag. PL Philip. (1904) 
111. 

Granum moschatum agreste Rumph. Herb. Amb. 4: 39. 
Amboina, Way tommo, Robinson PL Rumph. Amb. 493, August 17, 1913, 
along river banks, altitude about 50 meters. Erect, about 1.5 meters high. 

Granum moschatum agreste is very briefly described, but the 
probabilities are that the plant here determined as Abelmoschus 
mindanaensis Warb. is the one intended by Rumphius. The 
description given by Rumphius is mainly comparative with 
Abelmoschus moschatus Medik., a- taller, more woody plant than 
the latter, with which data the cited specimen agrees. It cannot 
possibly be Abelmoschus ficulneus W. & A. as suggested by 
Hasskarl, Neue Schlussel (1866) 74. The Amboina specimen is 
apparently a perfect match for our Mindanao material that un- 
questionably represents Warburg's species; it has previously 
been reported only from Mindanao. 

HIBISCUS Linnaeus 

HIBISCUS TILIACEUS Linn. Sp. PL (1753) 694. 

Paritium tiliaceum A. St. Hil. PL Bras. Merid. 1 (1825) 256. 

Novella Rumph. Herb. Amb. 2: 218, t. 73. 

Novella repens Rumph. Herb. Amb. 2: 222. 

Novella rubra Rumph. Herb. Amb. 2: 223. 
Amboina, Kati-kati, Robinson PL Rumph. Amb. 497, October 17, 1913, 
in cultivated ground, altitude about 80 meters, locally known as baru. 



MALVACEAE 



359 



Novella of Rumphius was first reduced to Hibiscus tiliaceus 
Linn, by Linnaeus, in Stickman Herb. Amb. (1754) 10, Amoen. 
Acad. 4 (1759) 121, Syst. ed. 10 (1759) 1149, Sp. PL ed. 2 
(1763) 976, which disposition of it is certainly correct and which 
has been accepted by practically all authors. The figure is not 
good. The form described by Rumphius as Novella reperis is 
certainly the form of Hibiscus tiliaceus Linn, with procumbent 
trunks that is very abundant in some localities, while Novella 
rubra is also manifestly merely a form of the same species. 

HIBISCUS M UTABI LIS Linn. Sp. PI. (1753) 694. 
Flos horarius Rumph. Herb. Amb. 4: 27, t. 9. 

This widely cultivated form is not represented in our Am- 
boina collections. The figure, however, is unmistakably refer- 
able to Hibiscus mutabilis Linn, and was first reduced to this 
species by Linnaeus himself, in Stickman Herb. Amb. (1754) 
15, Amoen. Acad. 4 (1759) 125, Syst. ed. 10 (1759) 1149, Sp. 
PL ed. 2 (1763) 977, in which he has been followed by all authors. 

HIBISCUS SURATTENSIS Linn. Sp. PI. (1753) 696. 

Hibiscus convolvulaceus Hassk. in Abh. Naurf. Gesellsch. Halle 9 

(1866) 216 (Neue Schliissel 74) (type!). 
Herba crinalium domestica Rumph. Herb. Amb. 4: 40, t. 16. 
Herba crinalium silvestris Rumph. 1. c. 41. 

Amboina, near the town of Amboina, Robinson PL Rumph.~ Amb. U96, 
August 8, 1913, near the beach. 

The original reduction of Herba crinalium to Hibiscus surat- 
tensis Linn, was made by Linnaeus, in Stickman Herb. Amb. 
(1754) 15, Amoen. Acad. 4 (1759) 126, Syst. ed. 10 (1759) 1145, 
Sp. PL ed. 2 (1763) 979, which is certainly the correct disposition 
of it. This reduction has been followed by all authors. Hass- 
karl, however, Neue Schliissel (1866) 74, decided that Herba 
crinalium silvestris represented a distinct species, which he 
named and described after Rumphius, as Hibiscus convolvulaceus, 
the publication of which has been overlooked by all authors and 
is not included in Index Kewensis. I consider this to be merely 
a form of Hibiscus surattensis Linn, with narrowly lobed leaves. 

HIBISCUS ROSA-SINENSIS Linn. Sp. PI. (1753) 694. 
Flos festal is Rumph. Herb. Amb. 4: 24, t. 8. 

This reduction was first made by Linnaeus, in Stickman Herb. 
Amb. (1754) 15, and the figure was consistently cited by Lin- 
naeus in his subsequent writings; while his reduction, certainly 
correct, has been followed by all authors. The form figured is 
the commonly cultivated one with double flowers. The form 



360 



RUMPHIUS'S HERBARIUM AMBOINENSE 



described as ruber simplex is the ordinary type with normal red 
flowers. Other forms described are the one with yellowish 
double flowers and one with normal, pale or nearly white flowers. 
This commonly cultivated plant does not occur in our Amboina 
collections, but is doubtless common in Amboina in cultivation, 
as it is in all parts of Malaya. 

THESPESIA Solander 

THESPESIA POPULNEA (Linn.) Soland. ex Corr. in Ann. Mus. Paris 9 
(1807) 290. 

Hibiscus populneus Linn. Sp. PI. (1753) 694. 
Thespesia macrophylla Blume Bijdr. (1825) 73, 106. 
Novella litorea Rumph. Herb. Amb. 2: 224, t. 7U. 

Amboina, Amahoesoe and Binting, Robinson PL Rumph. Amb. £98, 
August and September, 1913, along the strand. 

Novella litorea was first reduced by Linnaeus to Hibiscus 
populneus Linn., in Stickman Herb. Amb. (1754) 10, Amoen. 
Acad. 4 (1759) 121, Syst. ed. 10 (1759) 1149, Sp. PL ed. 2 (1763) 
976, which, as Thespesia populnea Soland., seems to be the correct 
disposition of it. Blume, however, Bijdr. (1825) 73, has pro- 
posed Thespesia macrophylla as a distinct species, apparently 
based wholly on the Rumphian figure and description, and many 
authors have recognized it as a valid species. Any large series 
of specimens from the Malayan region presents such relatively 
great variations in the characters by which the two forms have 
been distinguished, that I am by no means certain that they are 
specifically distinct and prefer to retain the Amboina plant under 
the older name.* 

GOSSYPIUM Linnaeus 

GOSSYPIUM BRASILIENSE Macf. PI. Jam. 1 (1837) 72. 

Gossypium lapideum Tussac, Fl. Antil. 2 (1818) 67, nomen nudum. 

Gossypium latifolium Rumph. Herb. Amb. 4: 37, t. 13. 
Amboina, Koesoe koesoe sereh, Robinson PI. Rumph. Amb. August 
23, 1913, locally known as kapas. 

The Rumphian species was first reduced by Linnaeus to Gos- 
sypium arbor eum Linn., in Stickman Herb. Amb. (1754) 15, 
Amoen. Acad. 4 (1759) 126, Syst. ed. 10 (1759) 1148, Sp. PL 
ed. 2 (1763) 975, where it certainly does not belong. Lamarck 
cites it in the original description of Gossypium vitifolium, 
Encycl. 2 (1786) 135, but Lamarck's type was an actual specimen 
collected by Sonnerat. Watt, Wild and Cultivated Cotton Plants 
of the World (1907) 255, leaves the Rumphian name as a syn- 

* See Baker in Journ. Bot. 35 (1897) 52. 



BOMBACACEAE 



361 



onym of Lamarck's species, but states that "Rumphius's Celebes 
plant shows the leaves too deeply 5-lobed (as in G. brasiliense) 
to be typical G. vitifolium." Other names involved in the reduc- 
tion are Gossypium nigrum Ham. and G. indicum Lam. While 
the disposition of Gossypium latifolium Rumph. as a synonym of 
G. brasiliense Macf. is not certainly the correct one, it is a 
reasonably safe one for the present. 

GOSSYPIUM INDICUM Lam. Encycl. 2 (1786) 134. 
Gossypium nanking Meyen Reise 2 (1836) 323. 
Gossypium Rumph. Herb. Amb. 4: 33, t. 12. 

This species is not represented in our Amboina collections. 
The Rumphian plant was first reduced by Linnaeus to Gossypium 
herbaceum Linn., in Stickman Herb. Amb. (1754) 15, Amoen. 
Acad. 4 (1759) 126, Syst. ed. 10 (1759) 1148, Sp. PI. ed. 2 
(1763) 975, where it certainly does not belong. Lamarck, 
Encycl. 2 (1786) 134, cited the Rumphian figure as representing 
his Gossypium indicum, although the actual type of that species 
is a specimen collected by Sonnerat. However, Watt considers 
that Gossypium indicum Lam. is a synonym of G. nanking Meyen, 
and as Lamarck's specific name is the oldest valid one for the 
species it is here accepted. Watt, Wild and Cultivated Cotton 
Plants of the World (1907) 128, definitely refers Rumphius's 
figure to Gossypium nanking var. nadam Watt. Hasskarl, Neue 
Schlussel (1866) 73, interprets the Rumphian description as 
Gossypium minus and G. majus, considering the tall form de- 
scribed to be Gossypium arboreum Linn. 

GOSSYPIUM PURPURASCENS Poir. in Lam. Encycl. Suppl. 2 (1811) 369. 
Gossypium floribus f usco-rubentibus Rumph. Herb. Amb. 4: 34. 

The reduction merely follows Hasskarl's suggestion. It is 
probable that it is the correct disposition of the Javan form 
that Rumphius casually and very briefly describes, but the data 
given are too few to warrant a certain identification of it. 

BOMBACACEAE 

DURIO Adanson 

DURIO ZIBETHINUS Murr. Syst. (1774) 591. 
Durio Rumph. Herb. Amb. 1: 99, t. 29. 

Amboina, Batoe gadjah, Robinson PI. Rumph. Amb. 69, August 5, 1913, 
from a cultivated tree, locally known as durion. 

Durio was reduced to Durio zibethinus by Murray in the orig- 
inal description of the species, the characters being apparently 



362 



RUMPHIUS'S HERBARIUM AMBOINENSE 



taken largely, if not wholly, from Rumphius. The three forms 
briefly described by Rumphius on page 101 may be merely 
variants of the common durian or some of them may represent 
distinct species. 

CEIBA Medikus 

CEIBA PENTANDRA (Linn.) .Gaertn. Fruct. 2 (1791) 244, t. 133. 
Bombax pentandrum Linn. Sp. PI. (1753) 511. 
Eriodendron anfractuosum DC. Prodr. 1 (1824) 479. 
Gossampinus alba Ham. in Trans. Linn. Soc. 15 (1826) 126. 
Gossampinus rumphii Schott. & Endl. Meletem. (1832) 35. 
Eriophorus javana Rumph. Herb. Amb. 1 : 194, t. 80. 

The common kapoc or silk cotton tree is not represented in 
our Amboina collections. Eriophorus javana Rumph. was first 
reduced by Linnaeus to Bombax pentandrum Linn., in Stickman 
Herb. Amb. (1754) 8, Amoen. Acad. 4 (1754) 120, Sp. PL ed. 
2 (1763) 958, but in the Systema, ed. 10 (1759) 1141, it is placed 
under Bombax aculeatum Linn., which may prove to be merely 
a synonym of Ceiba pentandra Gaertn. Bombax aculeatum 
Linn, does not appear in Index Kewensis. 

STERCULIACEAE 

PENTAPETES Linnaeus 

PENTAPETES PHOENICEA Linn. Sp. PL (1753) 698. 
Flos in pi us Rumph. Herb. Amb. 5: 288, t. 100, f. 1. 

This well-known species is not represented in our Amboina 
collections. The Rumphian figure and description are unmistak- 
ably the same as Pentapetes phoenicea Linn., and Linnaeus 
himself made the first reduction, in Stickman Herb. Amb. (1754) 
22, Amoen. Acad. 4 (1759) 131, Syst. ed. 10 (1759) 1150, Sp. 
PL ed. 2 (1763) 958. This reduction has been followed by all 
subsequent authors. 

COMMERSONIA Forster 

COM M ERSON I A BARTRAM I A (Linn.) comb. nov. 

Muntingia bartramia Linn. Amoen. Acad. 4 (1759) 124 (type!). 

Commersonia echinata Forst. Char. Gen. (1776) 44, t. 22. 

Restiaria alba Rumph. Herb. Amb. 3: 187, t. 119. 
Amboina, Hoenoet and Soja, Robinson PI. Rumph. Amb. 29U, August 2, 
1913, in flower, October 18, 1913, in fruit, locally known as marong, morong 
mera, and hunut. In clearings, altitude 200 to 375 meters. 

Restiaria alba Rumph. was reduced by Linnaeus, Amoen. Acad. 
4 (1759) 124, to Muntingia bartramia Linn, and is the whole 
basis of the Linnean species, which has been entirely overlooked. 



STERCULIACEAE 



363 



In the same year Linnaeus, Syst. ed. 10 (1859) 1044, erroneously 
referred the same figure to Triumfetta bartramia Linn, The 
Rumphian plate and description were reduced to Forster's species 
by Linnaeus f., Suppl. (1781) 187, and this disposition of it has 
been accepted by subsequent authors. In this connection botan- 
ists generally have recognized Commersonia echinata Forst. and 
C. platyphylla Andr. as two distinct species. The latter is hardly 
more than a form or a variety of the former. Gagnepain, Not. 
Syst. 1 (1909) 96, in a note regarding the typical form of Fors- 
ter's species, reduces Commersonia platyphylla Andr. to C. echi- 
nata Forst. var. platyphylla (Andr.) Gagnep. Kuntze, Rev. Gen. 
PI. 1 (1891) 81, recognizes Restiaria Rumph. as the proper 
generic name for Commersonia, but this is inadmissible under 
the rules of the International Code of Botanical Nomenclature. 

HELICTERES Linnaeus 

H ELICTERES ISORA Linn. Sp. PL (1753) 963. 

Fructus regis Rumph. Herb. Amb. 7: 32, t. 17, f. 1. 

This characteristic species is not represented in our Amboina 
collections. Burman, in the explanation of the Rumphian fig- 
ure, p. 33, connects Fructus regis with Helicteres isora Linn., 
citing not the Linnean binomial, but the diagnostic sentence 
"Helicteres foliis cordatis, serratis, fructu composito contorto, 
Linnaei Spec. Plant, pag. 963." This reduction was accepted by 
Linneaeus, Amoen. Acad. 4 (1759) 136, and manifestly is the 
correct disposition of Fructus regis. Lamarck, Encycl. 3 (1789) 
88, erroneously considered it to represent a variety of his Helic- 
teres ovata, a Brazilian species; while Hasskarl, PL Jav. Rar. 
(1848) 308, Neue Schlussel (1866) 189, adds Isora corylifolia 
Schott & Endl., a synonym of Helicteres isora Linn. 

KLEINHOVIA Linnaeus 
KLEI N HOV! A HOSPITA Linn. Sp. PI. ed. 2 (1763) 1365. 
Catti marus Rumph. Herb. Amb. 3: 177, t. 113. 
Amboina, Paso, Robinson PL Rumph. Amb. 292, September 8, 1913, 
margins of cultivated fields, altitude about 10 meters, locally known as 
kinar. 

Catti marus was cited by Linnaeus in the original description 
of Kleinhovia hospita, but the type was an actual specimen from 
Java, collected by Kleinhof; the generic name is corrected by 
some botanists to Kleinhofia, but the original Linnean spelling 
is here retained. All authors subsequent to Linnaeus who have 
cited the Rumphian description and figure have followed Lin- 
naeus in the reduction of Catti marus. 



364 



RUMPHIUS'S HERBARIUM AMBOINENSE 



STERCULIA Linnaeus 
STERCU LI A FOETIDA Linn. Sp. PL (1753) 1008. 

Clompanus molucanus Raf. Sylva Tellur. (1838) 73 (type!). 
Clompanus major Rumph. Herb. Amb. 3: 168, t. 107. 
Amboina, Silali, Robinson PI. Rumph. Amb. 293, in clearings at an 
altitude of about 150 meters, September 22, 1913. 

This reduction, manifestly correct, was first made by Linnaeus, 
in Stickman Herb. Amb. (1754) 14, Amoen Acad. 4 (1759) 
124, Syst. ed. 10 (1759) 1277, Sp. PL ed. 2 (1763) 1431, and 
has generally been accepted by all botanists who have had 
occasion to cite the Rumphian figure and description. The 
Rumphian description and figure typify Rafinesque's genus and 
species Clompanus molucanus. 

STERCULIA TREUBII Hochr. PI. Bogor. Exsicc. (1904) 8. 
Clompanus minor Rumph. Herb. Amb. 3: 169, t. 107 bis. 

Amboina, Lateri, Robinson PI. Rumph. Amb. 327, August 25, 1913; Ama- 
hoesoe, Robinson PL Rumph. Amb. 398,' August 28, 1913; the former in 
forests at an altitude of about 250 meters, the latter on coral rocks at an 
altitude of about 10 meters; locally known as choklat utan and saklat utan; 
that is, wild chocolate, the seeds being' used as a substitute for, or as an 
adulterant of, chocolate. 

Linnaeus reduced this to Sterculia balanghas Linn., in Stick- 
man Herb. Amb. (1754) 14, Amoen. Acad. 4 (1759) 124, Syst. 
ed. 10 (1759) 1277, Sp. PL ed. 2 (1763) 1430, and this disposi- 
tion of the Rumphian plant has been accepted by most authors. 
However, it manifestly is incorrect, as Sterculia balanghas Linn, 
is known only from India and Ceylon. Smith placed the Rumph- 
ian species under his Sterculia urceolata in the original de- 
scription of that species, Rees's Cyclop. 34 (1816) no. 3, in which 
he was followed by de Candolle, Prodr. 1 (1824) 482. Smith's 
material was from Honimoa Island, near Amboina, and his 
description, a copy of which has kindly been supplied to me by 
Sir David Prain, does not conform with my interpretation of 
Clompanus minor Rumph. The specimens cited above appear 
to be identical with Sterculia treubii Hochr., originally described 
from trees cultivated in the botanic garden at Buitenzorg, Java, 
their definite origin being unknown. The Rumphian figure is 
very poor, and from it alone one would hardly suspect the species 
described by Rumphius to be even closely allied to Sterculia 
treubii Hochr. The Rumphian description applies to the speci- 
mens much closer than the figure. 

Under Clompanus minor several forms are discussed, which 
probably represent distinct species. These are Clompanus terna- 
tensis femina Rumph. Herb. Amb. 3: 170; C. ternatensis mas Rumph. 



DILLENIACEAE 



365 



1. c, from Ternate; and C. silvestris Rumph. 1. c. 171, from Ceram. 
These are entirely undeterminable from the material and data 
now available, and their identity must await the results of field 
work in the two islands mentioned. 

ABROMA Jacquin 
ABROMA FASTUOSA Jacq. Hort. Vind". 3 (1776) 3, t. 1. 

Abroma (Ambroma) augusta Linn. f. Suppl. (1781) 341. 
Gossypium daemonis Rumph. Herb. Amb. 4: 38, t. 1^. 

No representative of the genus Abroma occurs in our Amboina 
collection, yet Gossypium daemonis Rumph. is manifestly an 
Abroma. Linnaeus, in Stickman Herb. Amb. (1754) 15, Amoen. 
Acad. 4 (1759) 126, erred in reducing it to his Hibiscus zeyla- 
nicus. Roxburgh, Fl. Ind. ed. 2, 3 (1832) 156, seems to be 
the first author to make the reduction to Abroma, by referring 
it to Abroma augusta Linn. f. The form described by Rumphius 
is manifestly what is currently named Abroma fastuosa R. Br.; 
that is, the form with spiny branchlets. Abroma mollis DC, 
Prodr. 1 (1824) 485, is described from specimens originating in 
the Moluccas, and is unquestionably the same as the spiny form 
currently named Abroma fastuosa R. Br. I have adopted the 
oldest name, Abroma fastuosa Jacq., but if this smooth-stemmed 
form be really specifically distinct from the form with the spiny 
stems, some adjustment of the synonymy will be necessary. 

H ERITI ERA Dryander 
H ERITI ERA LITTORAL IS Dryand. in Ait. Hort. Kew. 3 (1789) 546. 
Atunus litorea Rumph. Herb. Amb. 3: 95, t. 63. 

This common and widely distributed strand plant is not rep- 
resented in our Amboina collections. The reduction was first 
suggested by Lamarck, Encycl. 4 (1797) 228, and has generally 
been accepted by subsequent authors. Roxburgh, Fl. Ind. ed. 

2, 3 (1832) 142, erroneously reduced it to Heritiera minor Roxb., 
and by others it has been referred to Heritiera fomes Ham. 
It is, however, unquestionably Heritiera littoralis Dryand. 

DILLENIACEAE 

TETRACERA Linnaeus 
TETRACERA SCANDENS (Linn.) comb. nov. 

Tragia scandens Linn, in Stickman Herb. Amb. (1754) 18, Amoen. 

Acad. 4 (1759) 128 (type!). 
Delima hebecarpa DC. Syst. 1 (1818) 407. 

Tetracera hebecarpa Boerl. in Cat. Hort. Bot. Bogor. (1899) 3. 
Funis urens aspera Rumph. Herb. Amb. 5: 13, t. 9. 

This species is not represented in our Amboina collections, 



366 



RUMPHIUS'S HERBARIUM AMBOINENSE 



yet Funis urens asp era Rumph. is unmistakably a Tetracera 
and apparently the form described by de Candolle as Tetracera 
hebecarpa, which some authors reduce to Tetracera sarmentosa 
Linn., and others treat as a variety of the latter species. The 
Rumphian figure and description are the whole basis of Tragia 
scandens Linn., a species that -is not included in Index Kewensis. 
Linnaeus later, Sp. PI. ed. 2 (1763) 1390, referred Funis urens 
Rumph. to Tragia volubilis Linn., in which he was followed by 
Burman f., Poiret, and Henschel. Baillon, Etud. Gen. Euphorb. 
(1858) 461, 463, referred it to Tragia hirsuta Blume. The 
description, and for that matter the figure also, is unmistakably 
Tetracera and has nothing to do with the euphorbiaceous genus 
Tragia. 

TETRACERA BOERLAGEI sp. nov. 

Funis urens glabra Rumph. Herb. Amb. 5: 13. 
Amboina, Batoe gadjah, Robinson PL Rumph. Amb. 485 (type), August 
5, 1913, on trees at an altitude of about 200 meters. 

Frutex scandens, ramulis junioribus adpresse hirsutis; foliis 
oblongis ad oblongo-ellipticis, coriaceis, nitidis, usque ad 11 cm 
longis, glabris vel minutissime scaberulis, integris vel obscu- 
rissime undulatis, basi rotundatis ad subacutis, apice acutis 
ad obtusis, nervis utrinque 10 ad 12; inflorescentiis anguste 
pyramidatis, leviter hirsutis; sepalis interioribus obovatis, ro- 
tundatis, 6 mm longis, carpellis 2 vel 3, dense hirsutis ; f olliculis 
1 vel 2, rariter 3, oblongo-ovoideis, acuminatis, 8 ad 10 mm longis, 
parcissime hirsutis, in siccitate pallide olivaceis nitidisque, se- 
minibus solitariis. 

A scandent shrub, nearly glabrous except the very young 
branchlets and the inflorescence. Branches glabrous, reddish- 
brown, twisted, slender, tips of the young branchlets sparingly 
appressed-hirsute. Leaves oblong to oblong-elliptic, coriaceous, 
brittle, glabrous on both surfaces or very minutely and 
slightly scaberulous, in young leaves the midrib on the lower 
surface very slightly appressed-hirsute, 6 to 11 cm long, 2.5 
to 5.5 cm wide, rather pale or brownish when dry, shining, 
base rounded to subacute, apex acute to obtuse, margins entire 
or very obscurely undulate; lateral nerves 10 to 12 on each side 
of the midrib, prominent, brown; petioles 1 cm long or less. 
Panicles terminal, narrowly pyramidal, up to 12 cm in length, 
sparingly hirsute with scattered, subappressed hairs, the 
branches few, distant, the lower ones 2 cm long or less. Fruiting 
calyx with five sepals, the outer ones broadly ovate, slightly 
hirsute, about 3 mm long, the inner two or three obovate, 



DILLENIACEAE 



367 



rounded, about 6 mm long, their margins minutely ciliate. Car- 
pels 2 or 3, narrowly ovoid, densely hirsute, the styles glabrous. 
Follicles 1 or 2, sometimes 3, narrowly oblong-ovate, acuminate, 
8 to 10 mm long, pale olivaceous when dry, shining, with few, 
scattered, subappressed, rather long, stiff hairs, when fresh 
violet, turning brownish. Seeds solitary, broadly ovate, 2 to 
2.5 mm long, the aril pale, membranaceous, loose, obovate, 
rounded, 3 mm long, entire or obscurely toothed, not lacerate. 

This species is dedicated to Doctor J. G. Boerlage, who con- 
tracted a fever, while carrying on a botanical exploration of 
Amboina, which resulted in his death. It is manifestly allied to 
Tetracera indica Merr. (T. assa DC), from which it differs 
in its smaller sepals and follicles, the latter being prominently 
acuminate and sparingly hirsute. From the only other endemic 
Amboina species, Tetracera moluccana Martelli, it differs in its 
much shorter petioles and smaller fewer-nerved leaves. While 
Rumphius's description is short and largely comparative with 
Funis urens asp era, the specimen cited above agrees with it in 
all particulars, and I consider it certainly to represent the Rumph- 
ian plant. 

Ay-assa Rumph. Herb. Amb. 7 : 20 has erroneously been referred 
by some authors to Tetracera assa DC, but the plant that Rumph- 
ius describes presents little in common with Tetracera and is 
certainly not referable to this genus. Christmann and Panzer, 
Pflanzensyst. 4 (1779) 40, t. 26, f. 1, after Houttuyn, Nat. Hist. 
Plantenk. 5 (1776) 275, referred the Rumphian plant to Assa 
indica Christm. & Panz., but the description was manifestly based 
on an actual specimen. The generic name was apparently taken 
from Rumphius, and de Candolle, Syst. 1 (1818) 402, selected 
assa as the specific name under Tetracera. The synonymy should 
be adjusted, for Assa indica presents the oldest valid name: 
Tetracera indica (Christm. & Panz.) comb. nov. (Assa indica 
Christm. & Panz. Pflanzensyst. 4 (1779) 40, t. 26, f. 1; Tetracera 
assa DC Syst. 1 (1818) 402). 

DILLENIA Linnaeus 

DILLENIA ELLIPTICA Thunb. in Trans. Linn. Soc. 1 (1791) 200 (type!). 
Song i urn Rumph. Herb. Amb. 2: 140, t. 45. 

No representative of this species occurs in our Amboina col- 
lections. Dillenia elliptica Thunb. was based wholly on the 
Rumphian figure and description. Martelli, in Beccari Malesia 3 
(1887) 161, has redescribed Dillenia elliptica Thunb. from speci- 
mens collected in Celebes by Beccari, and doubtless this interpre- 



368 



RUMPHIUS'S HERBARIUM AMBOINENSE 



tation of the species is correct. Rumphius states that Songium 
was known in Amboina as aylassalinu and as ay macaninu, in 
Celebes as songi and songo, and in Java as sambu. Sempu is 
one of the Javanese names for Dillenia indica Linn., the only 
white-flowered Dillenia reported from Java, so that the Javan 
reference included by Rumphius probably refers to Dillenia 
indica Linn. Linnaeus reduced both t. J^5 and J>6 of Rumphius 
to Dillenia indica Linn., in Stickman Herb. Amb. (1754) 9, 
Amoen. Acad. 4 (1759) 121, Syst. ed. 10 (1759) 1082, but later 
authors have followed Thunberg in the disposition of t. 4-5. Lin- 
naeus was manifestly wrong in his reduction of t. U6, which is 
the following species. Miquel, Fl. Ind. Bat. I 2 (1858) 11, erro- 
neously reduced Dillenia elliptica Thunb. to D. speciosa Thunb., 
which is a synonym of D. indica Linn. 

DILLENIA SERRATA Thunb. in Trans. Linn. Soc. 1 (1791) 201 (type!). 
Sangius mas et femina Rumph. Herb. Amb. 2: 142, t. 4.6. 

This species is not represented in our Amboina collections. 
Dillenia serrata Thunb. was based wholly on Rumphius and 
must be interpreted from the Rumphian figure and description. 
Linnaeus erred in reducing this plate to his Dillenia indica, in 
Stickman Herb. Amb. (1754) 9, Amoen. Acad. 4 (1759) 121, 
Syst. ed. 10 (1759) 1082. All authors since Thunberg have 
been content with calling it Dillenia serrata, although the exact 
status of the species is quite uncertain. It seems probable that 
more than one species is included in the Rumphian description, 
as he gives the range as Amboina, Celebes, and Java. It was 
said by Rumphius to be abundant in Celebes at Tambocco, near 
Macassar, and at Toletae. Dillenia ochreata Teysm. & Binn. 
should be compared critically with Dillenia serrata Thunb., as 
it is probably identical with Thunberg's species. 

THEACEAE 

GORDONIA * Ellis 
GORDONIA RUMPHII sp. nov. 

Lignum muscosum 8. Caju lapia Rumph. Herb. Amb. 3: 203, t. ISO. 
Amboina, Hitoe messen, Robinson PI. Rumph. Amb. 276 (type), November 
1, 1913, in forests, altitude about 175 meters, locally known as kayu lapia. 

Arbor circiter 10 m alta, floribus exceptis glabra; foliis 
oblongis, coriaceis, usque ad 14 cm longis, acutis vel obscure 
acuminatis, obscure crenulatis, basi cuneatis, nervis utrinque 
circiter 15 ; floribus solitariis, axillaribus, circiter 3 cm diametro, 

* Retained name, Vienna Code; Lasianthus Adans. (1763) is older. 



THEACEAE 



369 



sepalis suborbicularibus, coriaceis, extus cinereo-pubescentibus, 
petalis extus filamentisque basi adpresse pubescentibus. 

A tree about 10 m high, quite glabrous except the inflorescence. 
Branches terete, reddish-brown, the branchlets smooth. Leaves 
oblong or oblong-elliptic, coriaceous, 8 to 14 cm long, 3.5 to 5.5 
cm wide, brown or brownish-olivaceous, prominently shining 
when dry, base cuneate, apex acute to obscurely acuminate, 
margins obscurely crenulate, the lower surface minutely verrucu- 
lose ; lateral nerves irregular, the primary ones about 15 on each 
side of the midrib, slender, anastomosing; petioles up to 5 mm 
in length. Flowers solitary, axillary, white, about 3 cm in 
diameter, their pedicels about 1 cm long, slightly pubescent. 
Sepals suborbicular, coriaceous, rounded, cinereous-pubescent ex- 
ternally, about 7 mm in diameter. Petals obovate, rounded, 
less than 1.5 cm long, externally pubescent. Stamens numerous, 
the basal portions of the filaments appressed-pubescent. Ovary 
ovoid, densely appressed-pubescent with pale hairs, the style- 
arms 5, short, glabrous. 

The identity of this plant with Lignum muscosum is certain. 
It bears the same native name as that cited by Rumphius for 
his plant, agrees well with his description, and fairly well with 
the rather crude plate. The identity of Lignum muscosum has 
not been previously determined, although Teysmann in a letter 
to Hasskarl considered it a species of Gordonia.* Its alliance 
is with Gordonia excelsa Blume, from which it is readily distin- 
guished by its entirely glabrous leaves and branchlets and its 
much smaller flowers. 

Caju lapia soyanansium s. lignum muscosum parvifolium Rumph. 
Herb. Amb. 3 : 203, mentioned and casually described under Caju 
lapia, is apparently an entirely different plant. Hasskarl, Neue 
Schliissel (1866) 67, mentions it, but suggests no identification of 
it. It is quite indeterminable from the data given by Rumphius. 

TERNSTROEMIA f Mutis 

TERNSTROE M I A ROBINSONII sp. nov. 

Ichthyoctonos montana Rumph. Herb. Amb. 3: 214, t. 139. 

Amboina, Hitoe messen, Robinson PI. Rumph. Amb. 275 (type) , October 
18, 1913, in forests, altitude about 350 meters locally known as anaan mera. 

Arbor circiter 14 m alta, glabra ; f oliis oblongo-obovatis, crasse 
coriaceis, usque ad 25 cm longis, petiolatis, apice obscure latis- 

* See Hasskarl, Neue Schliissel (1866) 67. 

f Retained name, Vienna Code; Mokof Adans. (1763), Taonabo Aubl. 
(1775), Dupinia Scop. (1777), and Hoferia Scop. (1777) are older. 

144971 24 



370 



RUMPHIUS'S HERBARIUM AMBOINENSE 



sime acuminatis ad obtusis, basi cuneatis, in siccitate purpureo- 
brunneis, nervis utrinque circiter 15; fructibus magnis, ellip- 
soideis, circiter 6 cm longis, calycis lobis valde incrassatis, 
sublignosis. 

A tree about 14 m high, quite glabrous, the branches stout, 
grayish. Leaves purplish-brown when dry, somewhat shining, 
thickly coriaceous, 15 to 25 cm long, 6 to 11 cm wide, generally 
oblong-obovate sometimes oblong, apex obtuse to very broadly 
and obscurely blunt-acuminate, base narrowed, cuneate; lateral 
nerves about 15 on each side of the midrib, slender, distinct, 
very obscurely anastomosing, the reticulations obsolete or nearly 
so; petioles stout, about 1 cm long. Flowers not seen. Fruits 
ellipsoid, vermilion when fresh, dark-brown when dry, about 
6 cm long, solitary, their pedicels stout, about 2 cm long, the 
calyx persistent, the lobes much thickened, rugose, somewhat 
woody when dry, more or less connate, the fruiting calyx 2 to 
2.4 cm in diameter, the pericarp smooth, rather brittle, each 
fruit containing three, red or garnet-colored, 3 cm long pyrenes 
which are notched at the apex, each containing two seeds. 

This species agrees sufficiently closely with Rumphius's 
description and with the rather crude figure. It is characterized 
by its relatively large leaves and large fruits and is most closely 
allied to Ternstroemia megacarpa Merr., of the Philippines ; from 
which, however, it differs in many characters, notably in its 
calyx-lobes being more or less connate, very much thickened, 
and somewhat woody when dry and in its much shorter pedun- 
cles. The form described by Rumphius in the same chapter as 
Ichthyoctonos litorea silvestris latifolia is possibly also referable here. 

GUTTIFERAE 

MESUA Linnaeus 

MESUA FERREA Linn. Sp. PI. (1753) 515. 

Calophyllum nagassarium Burm. f. PI. Ind. (1768) 121. 
Nagassarium Rumph. Herb. Amb. 7: 3, t. 2. 

This species is not represented in our Amboina collections. 
Nagassarium was described by Rumphius from specimens cul- 
tivated in Java. It was originally reduced to Mesua ferrea Linn, 
by Murray, Syst. (1774) 525, certainly the correct disposition 
of it, and one that has been accepted by practically all authors. 
Calophyllum nagassarium, Burm. f . is an exact synonym of Mesua 
ferrea Linn, and was based on Javan specimens with a reference 
to Nagassarium Rumph. as a synonym. 



GUTTIFERAE 



371 



CALOPHYLLUM Linnaeus 

CALOPHYLLUM INOPHYLLUM Linn. Sp. PI. (1753) 513. 

Balsamaria inophyllum Lour. Fl. Cochinch. (1790) 470. 
Calophyllum bintagor Roxb. Hort. Beng. (1814) 41 (type!), Fl. Ind. 

ed. 2, 2 (1832) 607. 
Bintangor maritima Rumph. Herb. Amb. 2: 211, t. 71. 
Amboina, Robinson PL Rumph. Amb. 480, September 22, 1913, along the 
seashore, locally known as bintangor. 

This is certainly the correct disposition of Bintangor marit- 
ima, a characteristic tree of tropical seashores of the Old World, 
the reduction having been made first by Linnaeus, in Stickman 
Herb. Amb. (1753) 10, Amoen. Acad. 4 (1759) 121, Syst. ed. 
10 (1759) 1075, Sp. PL ed. 2 (1762) 732, and generally accepted 
by all authors. It is the type of Calophyllum bintagor Roxb. as 
originally published in Hort. Beng. (1814) 41.* 

CALOPHYLLUM SOU LATTRI Burm. f. Fl. Ind. (1768) 121. 

Calophyllum spectabile Willd. in Ges. Naturf. Fr. Berl. Mag. 5 (1811) 
80. 

Bintangor silvestris Rumph. Herb. Amb. 2: 216, t. 72. 

Bintangor silvestris altera Rumph. 1. c. 217. 
Amboina, near houses, Robinson PI. Rumph. Amb. 482, August 20, 1913, 
locally known as sulatre; Hitoe messen, Robinson PL Rumph. Amb. 481, in 
forests, altitude 150 meters, locally known as bintangor utan. 

Bintangor silvestris was originally reduced by Linnaeus, 
through error, to the American Calophyllum calaba Linn., in 
• Stickman Herb. Amb. (1754) 10, Amoen. Acad. 4 (1759) 121. 
Lamarck, Encycl. 1 (1785) 553, cites it, with doubt, under 
Calophyllum acuminatum Lam., which is supposed to be a syn- 
onym of Calophyllum spectabile Willd. =C. soulattri Burm. f. 
Choisy, in de Candolle, Prodr. 1 (1824) 562, cites it as a doubtful 
synonym of Calophyllum spectabile Willd. ; while Hasskarl, Neue 
Schliissel (1866) 42, adds Calophyllum soulattri Burm. f., which 
is the oldest valid specific name for this widely distributed 
species. Rumphius's figure is poor, but his description conforms 
closely to the characters of the species. Bintangor silvestris 
altera Rumph. (PL Rumph. Amb. 481), seems to be merely a 
form of Calophyllum soulattri Burm. f . with the leaves somewhat 
retuse at the apex. 

CALOPHYLLUM sp. 

Bintangor montana (B. silvestris tertia) Rumph. Herb. Amb. 2: 217. 
Amboina, Hoetoemoeri road, Robinson PL Rumph. Amb. 483, September 
30, 1913, in forests, altitude about 300 meters. 

* See Robinson in Philip. Journ. Sci. 7 (1812) Bot. 414. 



372 



RUMPHIUS'S HERBARIUM AMBOINENSE 



The specimen certainly represents the plant that Rumphius 
briefly described as Bintangor montana, but it appears to be an 
undescribed species. Unfortunately no flowers are available, 
so that it is impossible to determine the true relationships of 
the form within the genus Calophyllum. The leaves are lanceo- 
late, 6 to 10 cm long, 2 to 3 cm wide, acute at the base, and 
gradually narrowed in the upper one-half to the rather slender 
and blunt-acuminate apex. The fruits are less than 1 cm in di- 
ameter. The plant is entirely glabrous except the ferruginous- 
pubescent buds. 

GARCINIA Linnaeus 

GARCINIA AM BOI N ENSIS Spreng. Syst. Veg. 2 (1825) 448 (type!). 
Folium acidum majus Rumph. Herb. Amb. 3: 58, t. 32. 

This is a species of very doubtful status, based entirely on 
Folium acidum majus Rumph. Loureiro, Fl. Cochinch. (1790) 
648, referred it to Oxy carpus cochinchinensis Lour. = Gar cinia 
cochinchinensis Choisy, a species based on Cochin-China speci- 
mens, quite different from the Amboina plant described by 
Rumphius and known only from Indo-China.* Doctor Robinson 
collected in Amboina typical Gar cinia dulcis Kurz, which he 
thought probably represented Folium acidum majus. However, 
while agreeing with the figure in many respects and with the 
description in part, the discrepancies are too great to warrant 
the citation of this specimen as representing the Rumphian plant. 
Boerlage, Cat. Hort. Bot. Bogor. (1899) 75, refers to Gar cinia 
amboinensis Spreng., two specimens, originating in Amboina and 
cultivated in the botanic garden at Buitenzorg. I have sterile 
specimens of one of these, "VI-F-11," which certainly agrees 
better with the original figure and description than does Gar- 
cinia dulcis Kurz, and which may represent Folium acidum majus 
in spite of certain discrepancies between the specimen and the 
figure and description. 

GARCINIA DULCIS (Roxb.) Kurz in Journ. As. Soc. Beng. 43 2 (1874) 88. 
Xanthochymus dulcis Roxb. PI. Coromandel 3 (1819) i. 270. ' 
Mundo Rumph. Herb. Amb. 1: 135. 
Amboina, Hitoe lama, Robinson PL Rumph. Amb. 478, October 8, 1913, in 
forests, altitude about 175 meters. 

The reduction of Mundo is based almost wholly on the native 
name, the brief description being entirely inadequate. It was 
from Java, and the native names mundo, munder, etc., are still 



* See Vesque in DC. Monog. Phan. 8 (1893) 449. 



GUTTIFERAE 



373 



in use there for this species. The type of Xanthochymus dulcis 
Roxb.= Gar cinia dulcis Kurz was from the Moluccas. 

GARCINIA CERAMICA Boerl. Cat. Hort. Bot. Bogor. (1899) 76? 
Folium acidum minus Rumph. Herb. Amb. 3: 60, t. 33. 

Folium acidum minus Rumph. is almost certainly a species 
of Garcinia, and the above reduction is suggested. There are, 
however, some discrepancies between the characters indicated 
by Rumphius and the authentic specimens of Boerlage's species 
before me. Vesque, in DC. Monog. Phan. 8 (1893) 349, has 
suggested that it may be the same as Garcinia picrorhiza Miq., 
but to me it seems much closer to G. ceramica Boerl. than to 
G. picrorhiza Miq. Two forms are described, majus and minus, 
which probably represent different species. Rumphius's ma- 
terial was not from Amboina, but from Little Ceram and Xula- 
bessi Islands. 

GARCINIA CAMBOGIA (Gaertn.) Desr. in Lam. Encycl. 3 (1791) 701. 
Mangostana cambogia Gaertn. Fruct. 2 (1791?) 106. 
Gutta cambodja Rumph. Herb. Amb. 2: 251. 

The form discussed by Rumphius as Gutta cambodja is un- 
doubtedly the same as Garcinia cambogia Desr., where it was 
placed by Hasskarl, Neue Schlussel (1866) 45. Henschel re- 
ferred it with doubt to Cambogia gutta Linn., but Cambogia 
gutta Linn, is a synonym of Garcinia morella Desr., not of Gar- 
cinia cambogia Desr.* 

GARCINIA CELEBICA Linn, in Stickman Herb. Amb. (1754) 7, Amoen. 
Acad. 4 (1759) 119, Syst. ed. 10 (1759) 1043, Sp. PI. ed. 2 (1762) 
635 (type!). 

Garcinia rumphii Pierre Fl. Forest. Cochinch. Enura. XIII, t. 77, A. 
Mangostana celebica Rumph. Herb. Amb. 1: 134, t. UU* 
Amboina, Way tommo, Robinson PI. Rumph. Amb. 323, August 16, 1913, 
on banks of the river at low altitudes. 

Garcinia celebica Linn, was based wholly on the Rumphian 
figure and description, which in turn were based on specimens 
from Macassar, Celebes, there known as kras and as mangostaan 
utan. Pierre, Fl. Forest Cochinch. Enum. XIII, not satisfied 
that specimens cultivated in the botanic garden at Buitenzorg, 
Java, under the name Garcinia celebica Linn, were correctly 
named, based his Garcinia rumphii on this material. I have 
a duplicate specimen of Pierre Ul 68, named Garcinia rumphii 

* See Trimen Fl. Ceyl. 1 (1893) 96. 



374 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Pierre, originating in the botanic garden at Buitenzorg, Java, 
which bears the native name kiras. Specimens of "VI-A-12a," 
"VI-A-16," and "VI-C-18a" from the Buitenzorg Botanic Garden 
certainly represent the same species, and agree well with the 
Rumphian figure and description of Mangostana celebica. Boer- 
lage, Cat. Hort. Bot. Bogor. (1899) 69, concluded also that 
Garcinia rumphii Pierre is the same as Mangostana celebica 
Rumph. and is a synonym of Garcinia celebica Linn. The Am- 
boina specimen collected by Robinson, cited above, should be 
critically compared with the closely allied Garcinia porrecta 
Wall. var. schizogyna Boerl. 1. c. 69. 

GARCINIA MANGOSTANA Linn. Sp. PI. (1753) 443. 
Mangostana Rumph. Herb. Amb. 1: 132, t. U3. 

The mangosteen is not represented in our Amboina collections. 
Mangostana was one of the few species figured and described 
by Rumphius that was reduced by Linnaeus in the first edition 
of the Species Plantarum, where it is cited under Garcinia 
mangostana Linn. The reduction is certainly correct and has 
been followed by all authors. 

GARCINIA CORNEA Murr. Syst. Veg. (1774) 368 (type!). 
Lignum corneum Rumph. Herb. Amb. 3: 55, t. 30. 

Not represented in our Amboina collections. Lignum cor- 
neum Rumph. is the whole basis of Garcinia cornea Murr., and 
the status of the species is now well known. It is a very charac- 
teristic species and occurs in cultivation in the botanic garden 
at Buitenzorg, Java; "VI-C-144" and "VI-C-144a" represent 
staminate and pistillate plants, both originating in Amboina. 
Murray has been consistently followed by all authors in this 
reduction of Lignum corneum. The two forms indicated and 
briefly described by Rumphius as latifolium and angustif olium 
are indeterminable from data now available. The former may 
be Garcinia latissima Miq., and the latter possibly Garcinia dulcis 
Roxb. 

GARCINIA PICRORHIZA Miq. Ann. Mus. Bot. Lugd. Bat. 1 (1866) 209. 
Pharmacum sagueri legitimum Rumph. Herb. Amb. 2: 136, t. UU- 

Boerlage, Cat. Hort. Bot. Bogor. (1899) 67, made this reduc- 
tion, and I consider that he is correct, after a comparison of 
the Rumphian figure and description with authentic specimens 
of Miquel's species, duplicates from the numbers cited by Boer- 
lage. Previous to Boerlage's reduction of the Rumphian plant 



DIPTEROCARPACEAE 



375 



to Garcinia picrorhiza Miq., its position had not been recognized 
by any author. 

GARCINIA PICRORHIZA Miq. var. LI MONORH IZA Boerl. Cat. Hort. Bot. 
Bogor. (1899) 68. 
Pharmacum limonicum Rumph. Herb. Amb. 2: 137, t. 4-4, f. B. 

This is, in all probability, correctly placed by Boerlage. It 
is in cultivation at Buitenzorg, but is not represented in Robin- 
son's Amboina collections. 

GARCINIA sp.? 

Vidoricum domesticum Rumph. Herb. Amb. 1 : 173. 

This form is of wholly doubtful status. Henschel thought 
that it might be a species of Bassia or of Diospyros. I have 
suggested Garcinia, as Rumphius states that the bark yields a 
yellow juice, and his description otherwise conforms fairly well 
with the characters of Garcinia. 

DIPTEROCARPACEAE 

SHORE A Roxburgh 

SHOREA SELANICA Blume Mus. Bot. 2 (1852) 33. 

Unona ? selanica DC. Prodr. 1 (1824) 92 (type!). 
Englehardtia selanica Blume Fl. Jav. 2 (1836) Jugl. 8 (type!). 
Hopea selanica W. & A. Prodr. (1834) 85; Walp. Repert. 5 (1845) 128 
(type!). 

Dammara selanica Rumph. Herb. Amb. 2: 168, t. 56. 

This species is not represented in our Amboina collections 
and is not credited to Amboina by Rumphius in the original 
description. The Rumphian description and figure are the 
whole basis of Unona selanica DC, Englehardtia selanica Blume, 
and Hopea selanica W. & A., and it is to be noted that Englehard- 
tia selanica Blume and Shorea selanica Blume were published 
without reference to the earlier Unona selanica DC. In trans- 
ferring the species to Shorea, Blume adds a short description, 
probably from Moluccan specimens. Burck, Ann. Jard. Bot. 
Buitenz. 6 (1887) 216, gives a more ample description based on 
specimens collected by Reinwardt and by Teysmann and on 
plants cultivated in the botanic garden at Buitenzorg, Java. 

SHOREA SELANICA Blume var. LATI FOLIA Blume Mus. Bot. 2 (1852) 33. 
Dammara selanica femina Rumph. Herb. Amb. 2: 169 (type!). 

Blume originally reduced Dammara selanica femina, by error, 
to Englehardtia spicata Blume Fl. Jav. 2 (1836) Jugl. 8, followed 
by Hasskarl's equally erroneous reduction of it to Englehardtia 



376 



RUMPHIUS'S HERBARIUM AMBOINENSE 



acerifolia Blume. Later Blume made it the type of Shorea 
selanica Bl. var. latifolia BL, which is perhaps the correct dis- 
position of it. The forms described in this chapter as Caju cawan 
e Java and Dammar leomelaena are undeterminable; the latter is 
probably a species of Canarium. 

DRYOBALANOPS Gaertner 

DRYOBALANOPS AROMATICA Gaertn. Fruct. 3 (1805) 49, t. 186. 

Arbor camphorifera II occidentals Rumph. Herb. Amb. 7: 65, 68. 

The general discussion of camphor includes the true camphor, 
Cinnamomum camphora T. Nees & Eberm., as well as that 
produced in Malaya, the resin of Dryobalanops aromatica 
Gaertn. This reduction was made by Blume, Mus. Bot. 2 (1851) 
38, and by de Vriese, who placed it under Dryobalanops cam- 
phora Colebr., a synonym of D. aromatica Gaertn. 

DIPTEROCARPUS Gaertner 

DIPTEROCARPUS sp.? 

Arbor koring Rumph. Herb. Amb. 2: 74. 

The reduction of Arbor koring to Dipterocarpus is based 
wholly on the observation made by Hamilton that the oil produced 
by the tree was secured by the same method as that used in 
gathering the oil of Dipterocarpus. The probabilities are very 
great that this is the correct disposition of Arbor koring. No 
further reduction of it is possible from the data given by 
Rumphius. 

BIXACEAE 

BIX A Linnaeus 

BIXA ORELLANA Linn. Sp. PI. (1753) 512. 

Pigmentaria Rumph. Herb. Amb. 2: 79, t. 19. 

Pigmentaria Rumph. was first reduced to Bixa orellana Linn, 
by Linnaeus, in Stickman Herb. Amb. (1754) 9, Amoen. Acad. 
4 (1759) 120, Syst. ed. 10 (1759) 1074, which is manifestly the 
correct disposition of it and has been accepted by all authors. 

FLACOURTIACEAE 

PANGIUM Reinwardt 
PANGIUM EDULE Reinw. in Syll. Ratisb. 2 (1828) 12. 
Pangium Rumph. Herb. Amb. 2: 182, t. 59. 
Amboina, Kati-kati, Robinson PI. Rumph. Amb. 212, October 20, 1913, 
in light forests, altitude about 75 meters, locally known as pangi. 



FLACOURTIACEAE 



377 



The reduction of Pangium made by Reinwardt in the original 
publication of the genus and species has been followed by all 
authors and is manifestly the correct disposition of it. The 
species is of very wide distribution in the Malayan region. 

FLACOURTIA L'Heritier 

FLACOURTIA INDICA (Burm. f.) comb. nov. 

Gmelina indica Burm. f. Fl. Ind. (1768) 132, t. 39, f. 5. 

Mespilus sylvestris Burm. Index Univ. Herb. Amb. 7 (1755) [18] 

(type!), non Burm. 1. c. [14]. 
Flacourtia sepiaria Roxb. PI. Coromandel 1 (1795) 48, t. 68. 
Flacourtia ramontchi L'Herit. Stirp. Nov. (1784-85) 59, t. 30, 31. 
Spina spinarum I mas Rumph. Herb. T^mb. 7: 36, t. 19, f. 1, 2. 
Spina spinarum il femina Rumph. 1. c. 37. 

This species ' is not represented in our Amboina collections. 
Rumphius states, however, that the plant was an introduced 
one there, originating in Java, where it was common. Spina 
spinarum Rumph. is the whole basis of Mespilus sylvestris 
Burm., as published on page 18 of his Index Universalis; it is 
not included in Index Kewensis. The name is invalid, however, 
because Burman published the same binomial on page 14 of the 
same work for an entirely different species, Carissa carandas 
Linn, (see p. 425). I consider that the form figured and 
described by Rumphius is the same as Flacourtia sepiaria Roxb., 
from which I cannot distinguish F. ramontchi L'Herit. Lin- 
naeus cites the first figure as a synonym of Carissa spinarum 
Linn., but the plant actually described and hence the type of 
the species is a true Carissa; figure 3 of the same plate, the 
type of Mespilus silvestris Burm. Index Universalis [14] non 
[18], is apparently a true Carissa. Linnaeus, in his erroneous 
reduction of Spina spinarum Rumph., was followed by Murray, 
Lamarck, Willdenow, Roemer and Schultes, Dietrich, and 
Pritzel. Loureiro, Fl. Cochinch. (1790) 634, cites the Rumphian 
species under Stigmarota jangomas Lour. = Flacourtia jangomas 
(Lour.) Steud. By other authors it has been referred to 
Damnacanthus indicus Gaertn., of the Rubiaceae; to Flacourtia 
jangomas Steud.; to Roumea sp.= Flacourtia; and to Flacourtia 
cataphracta Roxb. It is possible that Spina spinarum II femina 
Rumph. represents a species distinct from Spina spinarum I mas. 
Burman's Gmelina indica supplies the oldest valid specific name 
for the species and is here adopted. Burman's type was from 
Java, for which he cites the Javanese name doery roekan. 



378 



RUMPHIUS'S HERBARIUM AMBOINENSE 



HOMALIUM Jacquin 

HOMALIUM FOETIDUM (Roxb.) Benth. in Journ. Linn. Soc. Bot. 4 

(1860) 37. 

Ludia foetida Roxb. Hort. Beng. (1814) 38, nomen nudum, Fl. Ind. 

ed. 2, 2 (1832) 508. 
Blackwellia foetida Wall. Cat. (1831) no. 4899, nomen nudum, Deles- 

sert Ic. 3 (1837) 32, t. 53. 
Blackwellia moluccana Blume Mus. Bot. 2 (1852) 27 (type!). 
Metrosideros molucca mas Rumph. Herb. Amb. 3: 25, t. 11. 

Nothing resembling this species is represented in our Amboina 
collections. Homalium foetidum Benth. includes more than one 
species, the Mergui specimen cited being apparently Homalium 
griffithianum Kurz. It is to be noted that Bentham does not 
cite Ludia foetida Roxb. as a synonym of Homalium foetidum. 
Benth. He based his species on Blackwellia foetida Wall., but 
Wallich quotes Ludia foetida Roxb. as a synonym; the type of 
Ludia foetida Roxb. was a specimen cultivated in Calcutta, orig- 
inating in Amboina. In Index Kewensis Ludia foetida Roxb. 
is reduced to Flacourtia sumatrana Planch. Blackwellia moluc- 
cana Blume is based wholly on the Rumphian description and 
figure. Teysmann, quoted by Hasskarl, Neue Schliissel (1866) 
47, thought that Metrosideros molucca mas might be a Helicia, 
but this suggested reduction is manifestly entirely wrong. 

CARICACEAE 

CARICA Linnaeus 

CARICA PAPAYA Linn. Sp. PI. (1753) 1036. 
Papaya vulgaris Lam. Encycl. 5 (1804) 2. 
Papaja mas et femina Rumph. Herb. Amb. 1: 145, t. 50, 51. 

Rumphius's illustrations of the common papaya are excellent. 
The first reduction to Carica papaya Linn, was made by Lin- 
naeus, in Stickman Herb. Amb. (1754) 7, Amoen. Acad. 4 (1759) 
119, which is manifestly the correct disposition of Papaja 
Rumph. and is generally accepted by all authors. 

DATISCACEAE 

OCTOMELES Miquel 

OCTOMELES SUMATRANA Miq. Fl. Ind. Bat. Suppl. (1860) 336. 

Octomeles moluccana Teysm. & Binn. ex Hassk. in Abhandl. Nuturf. 

Gesellsch. Halle 9 (1866) 208 (Neue Schliissel 66) (type!). 
Octomeles moluccana Warb. in Engl. Bot. Jahrb. 13 (1891) 386. 
Palacca Rumph. Herb. Amb. 3: 195, t. 125. 

Teysmann and Binnnendyck, quoted by Hasskarl, Neue 



BEGONIACEAE 



379 



Schliissel (1866) 66, correctly reduced Palacca Rumph. to Octo- 
meles, but considered the species to be distinct from O. suma- 
trana Miq., calling it Octomeles moluccana. Warburg later 
considered that the oriental Malayan form was distinct from 
the one found in the Sunda Islands and described it as new 
under Octomeles moluccana Warb. I believe that K. Schumann 
and Lauterbach were correct in reducing Octomeles moluccana 
Warb. to 0. sumatrana Miq. At any. rate, if two species are 
represented, Teysmann and Binnendyck should be quoted as the 
authority for Octomeles moluccana, for although they published 
no description, their name is typified by Palacca as described and 
figured by Rumphius. 

BEGONIACEAE 

BEGONIA Linnaeus 

BEGONIA TUBEROSA Lam. Encycl. 1 (1785) 393 (type!), excl. syn. B. 
capensis Linn. 

Diploclinium tuberosum Miq. Fl. Ind. Bat. 1 1 (1856) 685. 
Empetrum acetosum I album Rumph. Herb. Amb. 5: 457, t. 169, f. 2. 

Amboina, Batoe gadjah and Batoe merah, Robinson PL Rumph. Amb. 65, 
August, 1913, altitude 15 to 200 meters, on rocks. 

Empetrum acetosum was originally reduced by Linnaeus to 
Begonia obliqua Linn., in Amoen. Acad. 4 (1759) 133, Sp. PL ed. 
2 (1763) 1497, in which he was followed by Burman f., FL Ind. 
(1768) 222. This disposition of it was entirely erroneous, as 
Begonia obliqua Linn, is an American species. Lamarck, 
Encycl. 1 (1785) 393, made it the type of his Begonia tuberosa. 
It is true that Lamarck erroneously gives as the first citation 
Begonia capensis Linn., but that he intended the Rumphian 
figure to typify his plant is manifest from his specific name. 
This was taken from the tuber-like lower part of the plant as 
shown in Rumphius's figure, which, however, was intended merely 
to represent a portion of the rock on which the plant grows. 
De Candolle, Prodr. 15 1 (1864) 323, mentions this tuber-like 
portion of the drawing thus : "Ex ic. Rumphii tuber 3 poll, 
crassum et habitus totus Cyclaminis." 

Possibly referable here is Robinson PL Rumph. Amb. 66 from 
Lateri and Koesoekoesoe sereh, August and September, 1913, 
on rocks, with smaller leaves than No. 65, cited above, in which 
the leaves are distinctly purplish when dry. Doctor Robinson 
thought this might represent Empetrum acetosum rubrum 
Rumph. Herb. Amb. 5 : 457, but Rumphius describes this form 
as having leaves a palm wide, which is not true of No. 66. 



380 



RUMPHIUS'S HERBARIUM AMBOINENSE 



I am unable definitely to place the forms very briefly described 
by Rumphius under Empetrum acetosum II rubrum and III cordatum. 
The former Hasskarl, Neue Schliissel (1866) 146, thought might 
be referable to Begonia (Diploclinium) rubrum Blume and the 
latter to Begonia mollis A. DC. It seems probable that Em- 
petrum acetosum II rubrum was merely a broad-leaved form of 
E. acetosum I album=Begonia tuber osa Lam., but E. acetosum 
III cordatum, described , as pilose, probably represents an en- 
tirely different species, not, however, represented in our Am- 
boina collections. 

CACTACEAE 

OPUNTIA Tournefort 

OPUNTIA sp. 

Ficus indica Rumph. Herb. Amb. 4: 89. 

The description is very imperfect, but probably applies to 
Opuntia. Hasskarl, Neue Schliissel (1866) 80, thought that 
it might be Opuntia dillenii Haw. 

THYMELAEACEAE 

GYRINOPSIS Decaisne 

GYRINOPSIS BRACHYANTHA Merr. in Philip. Journ. Sci. 7 (1912) 
Bot. 313. 

Cortex filarius Rumph. Herb. Amb. 7: 13. 

Amboina, Way uri, Robinson PL Rumph. Amb. 27U, September 9, 1913, 
in forests, altitude about 100 meters, locally known as melowassi. 

The specimen of melowassi is in fruit, but is apparently iden- 
tical with the Luzon species described by me as Gyrinopsis 
brachyantha. It differs from G. cumingiana Dene, not only in 
its shorter flowers, but also in the venation of its leaves, all the 
veins being slender and indistinct, the primary not more prom- 
inent than the secondary ones. Hasskarl, Neue Schliissel 
(1866) 186, cites Cortex filarius as a synonym of Anassera mo- 
luccana Pers. and of A. rumphii Span.; but Persoon, Syn. 1 
(1805) 265, and Lamarck before him, under Anasser moluccana 
Lam., 111. 2 (1797) 40, cites not Cortex filarius Rumph. but Cortex 
foetidus Rumph. Herb. Amb. 7: 12, t. 7, which is a Pittosporum. 
The correct status of Cortex filarius Rumph. has not been 
previously indicated. The genus Gyrinopsis has been reported 
only from the Philippines, but apparently also occurs in Borneo. 



LYTHRACEAE 



381 



AQUILARIA* Lamarck 

AQUILARIA MALACCENSIS Lam. Encycl. 1 (1783) 49. 

Aquilaria secundaria DC. Prodr. 2 (1825) 59 (type!). 
Aquilaria ovata Cav. Diss. (1790) 377, t. 22J>. 

Agallochum secundarium (coinamense et malaicense) Rumph. Herb. 
Amb. 2: 34, 35, t. 10. 

This species was not described from Amboina material. It 
is apparently the same as Aquilaria malaccensis Lam. Lamarck 
cites the Rumphian description and figure in his original descrip- 
tion. The figure is cited by Loureiro, Fl. Cochinch. (1790) 267, 
under Aloexylum agallochum Lour., but with a reference to 
Agallochum Rumph. Herb. Amb. 2: 29= Aquilaria agallocha 
Roxb., not to Aquilaria secundarium Rumph. Roxburgh's Aqui- 
laria agallocha. is a species published quite independently of 
Aloexylum agallochum Lour, and is hence not to be interpreted 
by Loureiro's description. From the characters assigned by 
Loureiro to Aloexylum, his plant seems to belong in the Legu- 
minosae, although it has very generally been considered to be 
the same as Aquilaria agallocha Roxb. Aquilaria secundaria 
DC. was based wholly on Rumphius's description and figure of 
Agallochum secundarium and is a synonym of Aquilaria malac- 
censis Lam. 

AQUILARIA AGALLOCHA Roxb. Hort. Beng. (1814) 33, nomen nudum, 
DC. Prodr. 2 (1825) 59; Roxb. Fl. Ind. ed. 2, 2 (1832) 422. 
Agallochum s. Calambac Rumph. Herb. Amb. 2: 29. 

The status of Agallochum or Calambac is doubtful. It was 
not described from Amboina material, but probably is the same 
as the Indian Aquilaria agallocha Roxb. Agallochum "officina- 
rum Lam.," cited as Encycl. 1 (1783) 48 and listed in Index 
Kewensis as a synonym of Aquilaria malaccensis Lam. 1. c. 49, 
I consider has no status, as Lamarck certainly did not intend a 
publication, but merely discussed the plant under Bauhin's name, 
Agallochum officinarum Bauh. Pin. (1623) 393. 

LYTHRACEAE 

LAGERSTROEM I A Linnaeus 

LAGERSTROEM I A INDICA Linn. Syst. ed. 10 (1759) 1076 (type!), Sp. 
PL ed. 2 (1763) 734. 
Lagerstroemia chinensis Linn. Amoen. Acad. 4 (1759) 137 (type!). 
Tsjinkin Rumph. Herb. Amb. 7: 61, t. 28, /. 1. 

This species is not represented in our Amboina collections. 



* Retained name, Vienna Code; Agallochum Lam. (1783) is older. 



382 



RUMPHIUS'S HERBARIUM AMBOINENSE 



The plate is a fair representation of the common and well-known 
Lagerstroemia indica Linn. It is the whole basis of Lager- 
stroemia indica Linn, and of L. chinensis Linn, as originally 
published, both in the year 1759. As to priority of publication I 
have no means of determining between volume four of the Amoe- 
nitates Academicae and the tenth edition of the Systema, but 
as Linnaeus himself abandons the name Lagerstroemia chinensis 
in favor of L. indica and as L. indica Linn, is the name univer- 
sally used for this well-known species, it should be maintained. 
Lagerstroemia chinensis Linn, does not appear in Index Kewen- 
sis, but Lagerstroemia chinensis Lam. Encycl. 3 (1791) 375, also 
typified by Tsjinkin of Rumphius, is listed there as a synonym of 
L. indica Linn. 

LAWSONIA Linnaeus 

LAWSONIA INERMIS Linn. Sp. PI. (1753) 349. 
Cyprus Rumph. Herb. Amb. 4: 42, t. 17. 

The common henna is not represented in our Amboina col- 
lections. Cyprus was originally reduced to Lawsonia spinosa 
Linn, by Linnaeus, in Stickman Herb. Amb. (1754) 15, Amoen. 
Acad. 4 (1753) 126, Sp. PL ed. 2 (1762) 498, which is a synonym 
of Lawsonia inermis Linn. Some authors have referred Cyprus 
to Lawsonia alba Lam., L. inermis var. spinosa Pers., and L. 
alba var. spinosa Lam., but these are all synonyms of the common 
and widely distributed Lawsonia inermis Linn. 

PEMPHIS Forster 

PEMPHIS ACIDULA Forst. Char. Gen. (1776) 68, t. 34. 
Ly thrum pemphis Linn. f. Suppl. (1781) 249. 

Aegiceras ferreum Blume Bijdr. (1825) 693 p. p. quoad syn. Rumph. 
Mangium ferreum mas Rumph. Herb. Amb. 3: 120, t. 79, excl. f. A, B. 
Mangium porcellanicum Rumph. Herb. Amb. 3: 126, t. 84. 

This common and widely distributed strand plant is not rep- 
resented in our Amboina collections. Mangium ferreum mas 
as figured by Rumphius presents a flowering branch of Pemphis 
acidula Forst., but the additional figures A and B are Aegiceras. 
This mixture of the two species was first pointed out by Teys- 
mann, as quoted by Hasskarl, Neue Schliissel (1866) 57. The 
form described by Rumphius, 1. c, as Mangium ferreum femina 
is probably merely Pemphis acidula Forst. Mangium porcellani- 
cum Rumph. was first reduced to Lythrum pemphis Linn, by 
Retzius, Obs. 5 (1789) 4, and as Pemphis acidula Forst. this is 
the correct disposition of it. 



SONNERATIACEAE 



383 



SONNERATIACEAE 

SONNERATIA * Linnaeus f. 

SONNERATIA ALBA Sm. in Rees. Cycl. 33 (1816) no. 2. 

Mangium caseolare album Rumph. Herb. Amb. 3: 111, t. 73. 
Amboina, Wakeroe, and at Ayer putri, Robinson PI. Rumph. Amb. 290, 
July 28 and October 17, 1913, along tidal streams, locally known as mangi 
mangi. "Flower apetalous, sepals lilac inside." 

Mangium caseolare album was originally referred by Linnaeus 
to Rhizophora caseolaris, in Stickman Herb. Amb. (1754) 13, 
Linnaeus overlooking the fact that at least two distinct species 
were considered by Rumphius, one having flowers without petals, 
and one with petals. I have not seen the original description of 
Sonneratia alba Smith, which may have been based in part 
on Rumphius. Succeeding authors, de Candolle, Don, Blume, 
and Miquel, cite the Rumphian plate as representing Sonneratia 
alba Smith. 

SONNERATIA CASEOLARIS (Linn.) Engl, in Engl. & Prantl Nat. Pflanz- 
enfam. Nachtr. 1 (1897) 261. 
Rhizophora caseolaris Linn. p. p., in Stickman Herb. Amb. (1754) 13, 
Amoen. Acad. 4 (1759) 123, Syst. ed. 10 (1759) 1043, Sp. PI. ed. 2 
(1763) 635 (type!). 
Sonneratia acida Linn. f. Suppl. (1781) 252. 
Sonneratia pagatpat Blanco Fl. Filip. (1837) 424. 
Mangium caseolare rubrum Rumph. Herb. Amb. 3: 112. t. 7U. 
Amboina, Paso, Robinson PI. Rumph. Amb. 291, October 29, 1913, at the 
back of a mangrove swamp, "petals present, lilac." 

This is Rhizophora caseolaris Linn, in part only. The orig- 
inal description in Stickman, Herb. Amb. (1754) 13, is as fol- 
lows: "73. 75. Mangium caseolare [=] Rhizophora caseolaris, 
foliis ovatis obtusis, floribus solitariis, fructibus orbiculatis de- 
pressis mucronatis." The figures refer to plates 73 to 75 of 
Rumphius, inclusive. As noted above t. 73 represents the ape- 
talous Mangium caseolare album Rumph. and is Sonneratia alba 
Smith; t. 7U, however, represents Mangium caseolare rubrum 
Rumph., definitely described as having petals. Perhaps on a 
very strict interpretation of types, t. 73 should represent the 
plant now known as Sonneratia caseolaris (Linn.) Engl., in 
which case the name Sonneratia caseolaris would have to be 
applied to the plant now known as Sonneratia alba Sm., and 
what is here interpreted as Sonneratia caseolaris would have 

* Retained name, Vienna Code; Blatti Adans. (1763) and Pagapate Sonn. 
(1776) are older. 



384 



RUMPHIUS'S HERBARIUM AMBOINENSE 



to be called Sonneratia pagatpat Blanco. Botanists very gen- 
erally, since the publication of Sonneratia acida Linn. f. in 1781, 
have referred to it Mangium caseolare rubrum of Rumphius, 
including tt. 7U, 75; but Sonneratia acida Linn. f. is merely a 
synonym of Sonneratia caseolaris (Linn.) Engl., as interpreted 
by Engler and as interpreted here. 

PUNICACEAE 

PUNIC A Linnaeus 
PUNICA GRANATUM Linn. Sp. PI. (1753) 472. 

Malum granatum Rumph. Herb. Amb. 2: 94, t. 2U, f. 1. 

The common pomegranate is not represented in our Amboina 
collections, although it is found in scattered cultivation through- 
out the Malayan region. Malum granatum was first reduced to 
Punica granatum Linn, by Linnaeus, in Stickman Herb. Amb. 
(1753) 9, Amoen. Acad. 4 (1759) 120, Syst. ed. 10 (1759) 1056, 
which is manifestly the correct disposition of it. 

LECYTHIDACEAE 

BARRINGTONIA * Forster 

BARRI NGTON I A ASIATICA (Linn.) Kurz in Journ. As. Soc. Beng. 45 1 
(1876) 131; 46 2 (1877) 70. 
Mammea asiatica Linn. Sp. PL (1753) 512. 

Barringtonia speciosa Forst. Char. Gen. (1776) 76 t. 38, f. A-C ; Linn. 

f. Suppl. (1781) 312. 
Agasta asiatica Miers in Trans Linn. Soc. Bot. 1 (1875) 61. 
Agasta indica Miers 1. c. 63. 
Butonica rumphiana Miers 1. c. 68. 
Butonica Rumph. Herb. Amb. 3: 179, t. UU. 
Amboina, Robinson PL Rumph. Amb. Jf.66, September 16, 1913, along the 
river near the town of Amboina. 

Butonica was first reduced to Barringtonia speciosa by the 
younger Linnaeus, Suppl. (1781) 312, which has been followed 
by all authors except Miers. The latter retains Barringtonia 
speciosa Forst. as the sole representative of the genus which he 
confined to Polynesia and removed the Indo-Malayan forms from 
Barringtonia as Agasta asiatica (Linn.) Miers and A. indica 
Miers. He has not been followed by subsequent authors, the 
general conclusions regarding Miers's proposed classification, 
in which I concur, being that his three species are all merely 
forms of the common and widely distributed strand plant, Bar- 
ringtonia asiatica (Linn.) Kurz. The type of the Linnean 
species, Mammea asiatica, was collected by Osbeck on a small 



* Retained name, Vienna Code; Huttum Adans. (1763) is older. 



LECYTHIDACEAE 



3£5 



island near the western end of Java. Butonica rumphiana Miers 
is certainly identical with Mammea asiatica IArm.=Barringtonia 
asiatica Kurz. 

BARRINGTONIA RACEMOSA (Linn.) Blume ex DC. Prodr. 3 (1828) 288; 
Roxb. Fl. Ind. ed. 2, 2 (1832) 634. 

Eugenia racemosa Linn. Sp. PI. (1753) 471. 

Stravadia alba Pers. Syn. 2 (1807) 30. 

Stravadium album DC. Prodr. 3 (1828) 289. 

Barringtonia alba Blume in Fl. des Serres I 7 (1851-52) 23. 

Barringtonia Alba Kostel. Allgem. Med. Pharm. Fl. 4 (1835) 1536. 

Stravadium rubrum DC. Prodr. 3 (1828) 289, p. p., quoad syn. Rumph. 

Butonica terrestris Miers in Trans. Linn. Soc. Bot. 1 (1875) 69. 

Barringtonia rubra Blume Fl. des Serres I 7 (1851-52) 23. 

Butonica rubra Miers in Trans. Linn. Soc. Bot. 1 (1875) 70. 

Barringtonia inclyta Miers in Trans. Linn. Soc. Bot. 1 (1875) 71. 

Butonica terrestris rubra Rumph. Herb. Amb. 3: 181 ! (t. 115?). 

Butonica terrestris alba Rumph. Herb. Amb. 3: 181, t. 116. 
Amboina, Paso, Robinson Rumph. Amb. 467, November 25, 1913, near 
the beach. 

This much-named species is widely distributed along and 
near the seashore from India to Malaya and Polynesia. Miers 
has attempted with little success to distinguish several species. 
As the proper authority for Barringtonia racemosa, I have 
selected Blume (1828) in preference to Roxburgh, as Roxburgh's 
original use of the name, Hort. Beng. (1814) 52, is a nomen 
nudum. 

In previous reductions of the Rumphian descriptions and 
figures, practically all authors have assumed that two species 
were involved. Butonica terrestris alba was originally reduced 
by Linnaeus to Eugenia racemosa Linn., in Stickman Herb. Amb. 
(1754) 14, Amoen. Acad. 4 (1759) 124, Syst. ed. 10 (1759) 1055, 
Sp. PI. ed. 2 (1762) 673; but other authors have reduced it to 
Barringtonia acutangula Gaertn., to Stravadia alba Pers., to 
Barringtonia alba Kostel., to Butonica alba Miers, etc. The 
description and the figure apply unmistakably to Barringtonia 
racemosa (Linn.) Blume as here interpreted. 

The description of Butonica terrestris rubra Rumph. is un- 
mistakably Barringtonia racemosa (Linn.) Blume, as here 
interpreted, but the figure, which is poor, may represent an- 
other species, possibly Barringtonia acutangula Gaertn. In 
Rumphius's description note: 

Folia * * * unum vel sesquipedem longi, immo longiora * * * 
palmam nempe lata, vel paulo latiora * * *. Flores ex longo, tenui, & 
dependente petiolo, binos pedes longo * * * huic viridia insident capita 
instar Olivarum tenerarum, per ilium laxe dispersa, quae sese aperiunt in 
bina vel terna crassa & concava petala [sepals] . 

144971 25 



386 



RUMPHIUS'S HERBARIUM AMBOINENSE 



The description otherwise applies to Barringtonia racemosa, 
not to B. acutangula Gaertn., to which it has been reduced by 
many authors. It was originally reduced by Linnaeus to 
Eugenia acutangula Linn., in Stickman Herb. Amb. (1754) 14, 
Amoen. Acad. 4 (1759) 124, Sp. PL ed. 2 (1762) 673; by 
Lamarck, Encycl. 3 (1789) 197, to Eugenia racemosa Linn.; 
by Loureiro, Fl. Cochinch. (1790) 410, to Meteorus coccineus 
Lour., which is a possible synonym of Barringtonia racemosa 
Blume ; and by other authors to Barringtonia acutangula Gaertn., 
to Stravadia rubra Pers., to Stravadium rubrum DC., to Barring- 
tonia rubra Blume, and finally by Miers to Butonica terrestris 
Miers. 

RHIZOPHORACEAE 

The Rhizophoraceae described and figured by Rumphius are 
obscure, and the actual status of the several species involved is 
susceptible of various interpretations. The species actually rep- 
resented in our Amboina collections are the forms commonly 
known as Bruguiera eriopetala W. & A., B. caryophylloides 
Blume, B. parviflora W, & A., and Rhizophora conjugata Linn. 
By a strict interpretation of types, following the principles of 
priority, most of these names must be discarded, Bruguiera 
eriopetala W. & A. becoming B. sexangula (Lour.) Spreng., 
B. caryophylloides Blume becoming B. cylindrica (Linn.) Blume, 
and Rhizophora conjugata auct., non Linn., becoming R. cande- 
laria DC. Species apparently described by Rumphius, but not 
included in the Amboina collections at present available for 
study, are apparently Ceriops tagal (Perr.) C. B. Rob. (C. 
candolleana Arn.) and Bruguiera conjugata (Linn.) Merr. (B. 
gymnorhiza Lam.). A number of species and synonyms must 
be interpreted wholly or in part from the Rumphian figures and 
descriptions. An attempt has here been made to select the 
earliest valid specific name in each case and to adjust the syn- 
onymy, but a future monograph of the group, based on very 
comprehensive collections, may modify some of these conclusions. 

CERIOPS Arnott 

CERIOPS TAGAL (Perr.) C. B. Rob. in Philip. Journ. Sci. 3 (1908) Bot. 306. 
Rhizophora tagal Perr. Mem. Soc. Linn. Paris 3 (1824) 138. 
Rhizophora timoriensis DC. Prodr. 3 (1828) 32. 
Ceriops candolleana Arn. in Ann. Nat. Hist. 1 (1838) 363. 
Mangium caryophylloides II parvifolium et III latifolium Rumph. 
Herb. Amb. 3: 119. 

No representative of the genus Ceriops occurs in our Amboina 



RHIZOPHORACEAE 



387 



collections. The suggested reduction of the two forms described 
under Mangium caryophylloides Rumph. is after Hasskarl, Neue 
Schliissel (1866) 57, but Ceriops tagal is an older name than C. 
candolleana Arn. Blume, Mus. Bot. 1 (1849) 143, suggests that 
Mangium caryophylloides II may be the same as Ceriops zippel- 
iana Blume, and that Mangium caryopylloides III may be the 
same as Ceriops forsteniana Blume, but both of Blume's species 
are apparently merely forms of the common and widely dis- 
tributed Ceriops tagal C. B. Rob. Blume also suggests that 
Mangium minus Rumph., Herb. Amb. 3: 106 quoad descr., excl. t. 
69, may be Ceriops zippeliana Blume, which may be the correct 
disposition of the description; the plate is considered under 
Bruguiera conjugata (Linn.) Merr. and under B. cylindrica 
(Linn.) Blume (p. 388). 

RHIZOPHORA Linnaeus 

RHIZOPHORA CANDELARIA DC. Prodr. 3 (1828) 32 (type!). 
Rhizophora conjugata auct. plur., non Linn. 
Rhizophora apiculata Blume Enura. 1 (1828) 91. 
Mangium candelarium Rumph. Herb. Amb. 3: 108, t. 71, 72. 

Amboina, near the town of Amboina, Robinson PL Rumph. Amb. 266, 
July 26, 1913, along tidal streams, locally known as mangi mangi. 

While both plates given by Rumphius are very crude, they 
unmistakably represent the form with short peduncles bearing 
usually two flowers. The description, however, may include also 
the allied Rhizophora mucronata Lam., which is distinguished 
by its longer peduncles and more numerous flowers. By Lin- 
naeus it was erroneously reduced to the American Rhizophora 
mangle, in Stickman Herb. Amb. (1754) 13, Amoen. Acad. 4 
(1759) 123, Syst. ed. 10 (1759) 1043, Sp. PI. ed. 2 (1763) 634, 
in which he was followed by Lamarck, Willdenow, Burman f., 
Roxburgh, and other authors. Blume, Mus. Bot. 1 (1849) 132, 
placed it under Rhizophora mucronata Lam. with the comment 
"descriptio satis bona, figurae minus rectus !," while Walpers, 
Repert. 2 (1843) 70, placed it under Rhizophora conjugata Linn. 
Rhizophora candelaria DC. is a composite species to be inter- 
preted from Mangium candelarium Rumph. and Pee-kandel 
Rheed. Hort. Malabar. 6 : t. SU, Rheede being cited first, Rumph- 
ius second; but as the specific name is manifestly taken from 
Rumphius, while the description given by de Candolle applies 
better to the form generally interpreted as Rhizophora conjugata 
auct. (non Linn.) than to R. mucronata Lam. Rheede's figure 
is considered to represent Rhizophora mucronata Lam. 



388 



RUMPHIUS'S HERBARIUM AMBOINENSE 



BRUGU I ERA Lamarck 

BRUGUIERA CONJUGATA (Linn.) Merr. in Philip. Journ. Sci. 9 (1914) 
Bot. 118. 

Rhizophora conjugata Linn. Sp. PI. (1753) 443, non aliorum! 
Rhizophora gymnorhiza Linn. 1. c. 

Bruguiera gymnorhiza Lam. 111. 2 (1797) t. 397, Encycl. 4 (1798) 696. 
Bruguiera rumphii Blume Mus. Bot. 1 (1849) 138. 
Rhizophora ? palun DC. Prodr. 3 (1828) 33. 

Bruguiera gymnorhiza Lam. var. palun Blume Mus. Bot. 1 (1849) 
136. 

Mangium celsum Rumph. Herb. Amb. 3: 102, t. 68. 

Mangium minus Rumph. Herb. Amb. 3: 1Q6, t. 69 excl. fl. et /. A, B. 

There is no specimen of this species in our Amboina collec- 
tions; and, although both of the figures cited above are crude, 
there is very little doubt that both are properly referable here. 
The reduction of Mangium celsum was first made by Lin- 
naeus, in Stickman Herb. Amb. (1754) 12, Amoen. Acad. 4 
(1759) 123, Syst. ed. 10 (1759) 1043 (Rhizophora gymnorhiza 
Linn.), in which he was followed by various other authors. 
Since the transfer of the species to Brugiera, the plate has very 
generally been cited under B. gymnorhiza Lam., although Blume 
placed it under his B. rumphii. Mangium minus Rumph. is 
manifestly a mixture, the leaves and fruits apparently being 
the same as those of B. conjugata (B. gymnorhiza), but the 
flowers and figures A and B certainly represent Bruguiera cylin- 
drica Blume, below. It is, excluding figures A and B, the whole 
basis for Rhizophora palun DC. Blume, Mus. Bot. 1 (1849) 143, 
suggests that the description, but not the plate, of Mangium 
minus s. palun may be Ceriops zippeliana Blume (see Ceriops 
tagal (Perr.) C. B. Rob. p. 386). It is to be noted that the two 
attached flowers on the plate of Mangium minus Rumph. Herb. 
Amb. 3: 106, t. 69, and figures A and B, were copied by Burman 
from Rheede's figure of Cari Candel, Hort. Malabar. 6 : t. 33, and 
have nothing to do with the plant actually figured and described 
by Rumphius; see under Bruguiera cylindrica (Linn.) Blume, 
below. 

BRUGUIERA CYLINDRICA (Linn.) Blume Enum. PL Jav. (1828) 93. 

Rhizophora cylindrica Linn. Sp. PL (1753) 443. 
Rhizophora caryophylloides Burm. f. FL Ind. (1768) 109. 
Bruguiera caryophylloides Blume Enum. PL Jav. (1828) 93. 
Kanilia caryophylloides Blume Mus. Bot. 1 (1849) 141. 
Mangium minus Rumph. Herb. Amb. 3: t. 69, pp. quoad fl. et /. A, B. 
Mangium caryophylloides Rumph. Herb. Amb. 3: 119, t. 78. 
Amboina, near the town of Amboina, Robinson PI. Rumph. Amb. 265, 
along tidal streams, July 26, 1913, locally known as tonki. 



RHIZOPHORACEAE 



389 



Rhizophora cylindrica Linn, must be interpreted solely by 
Cari-Candel Rheed. Hort. Malabar. 6 : 59, t. 33, which most 
authors agree represents the form described by Blume as Bru- 
guiera caryophylloides (Burm. f.) Blume. Mangium minus 
Rumph., as figured, is manifestly a composite species, the at- 
tached flowers and figures A and B having been copied by 
Burman in editing Rumphius's work from Cari-Candel Rheed. 
Hort. Malabar. 6: t. 33. The leafy branch and fruits, excluding 
the attached flowers and figures A and B, and the description 
for the most part are probably referable to Bruguiera gymnor- 
hiza Lam. (B. rheedi Blume) . It was reduced by Linnaeus to his 
Rhizophora cylindrica, in Stickman Herb. Amb. (1754) 12, 
Amoen. Acad. 4 (1759) 123, Syst. ed. 10 (1759) 1043, Sp. PI. 
ed. 2 (1762) 635, in which he was followed by Burman, La- 
marck, Willdenow, and other authors. Mangium caryophylloides 
Rumph. is the basis of Rhizophora caryophylloides Burm. f., 
Fl. Ind. (1768) 109, from which again Bruguiera caryophylloides 
Blume and Kanilia caryophylloides Blume must be interpreted. 
The figure is not good, but is undoubtedly referable here, while 
the description applies closely to the form generally named Bru- 
guiera caryophylloides Blume=Z?. cylindrica (Linn.) Blume. 

BRUGUIERA SEXANGULA (Lour.) Poir. in Lam. Encycl. Suppl. 4 (1816) 
262. 

Rhizophora sexangula Lour. Fl. Cochinch. (1790) 297. 
Bruguiera eriopetala W. & A. in Ann. Nat. Hist. 1 (1838) 368. 
Bruguiera gymnorhiza Blume Mus. Bot. 1 (1849) 136, non Lam. 
• Bruguiera cylindrica Blume 1. c. 137. 
Mangium digitatum Rumph. Herb. Amb. 3: 107, t. 70. 
Amboina, Ayer putri, Robinson PI. Rumph. Amb. 264, July 28, 1913, 
along tidal streams, locally known as mangi mangi. 

Mangium digitatum Rumph. has been reduced by some authors 
to Bruguiera gymnorhiza Lam. and by others to B. cylindrica 
Blume. It seems, however, to be the species commonly known as 
Bruguiera eriopetala W. & A., which I have here reduced to the 
older Bruguiera sexangula Poir. (Rhizophora sexangula Lour.) . 

The generic description given by Linnaeus, Gen. PL (1754) 
202, is unmistakably Rhizophora as at present understood, 
although the first two species cited in his Species Plantarum 
(1753) 443 are representatives of Bruguiera, the third a 
Kandelia, and the last two true Rhizophorae. The first species, 
Rhizophora conjugata Linn., has been misinterpreted by most 
authors since its publication, but it has page priority over 
Rhizophora gymnorhiza Linn. Trimen * states : 



* Fl. Ceyl. 2 (1894) 154. 



390 



RUMPHIUS'S HERBARIUM AMBOINENSE 



There is no specimen in Hermann's Herb., but his drawing is unmis- 
takably this species [Bruguiera gymnorhiza Lam.] and it is the whole 
foundation for Linnaeus's Rhizophora conjugata, which name has been since 
always applied to another plant, R. Candelaria DC, to which this bears a 
strong resemblance in foliage. 

If rules of priority be followed, the adoption of the name 
Bruguiera conjugata is unavoidable for this widely distributed 
Indo-Malayan species. 

COMBRETACEAE 

TERM I N ALIA * Linnaeus 

TERM I N ALIA CATAPPA Linn. Mant. 1 (1767) 128, 2 (1771) 519. 

Terminalia moluccana Lam. Encycl. 1 (1783) 349 (type!). 

Juglans catappa Lour. Fl. Cochinch. (1790) 573. 

Catappa domestica Rumph. Herb. Amb. 1 : 174, t. 68. 
Amboina, Hatiwe, Robinson PL Rumph. Amb. September 4, 1913, 

along the seashore, locally known as katappan. 

The Rumphian figure and description are, at least in part, the 
basis of Terminalia catappa Linn., as they were cited in the 
original publication of the species. They are also the whole 
basis of Terminalia moluccana Lam., cited above, and in part the 
basis of Juglans catappa Lour. The three forms described by 
Rumphius are probably all referable to Terminalia catappa Linn., 
which presents considerable variation in its fruit characters. 
Hasskarl, Neue Schlussel (1866) 22, has referred all of them to 
varietal forms, Catappa domestica to Terminalia catappa var. 
macrocarpa Hassk., C. silvestris litorea to T. catappa var. 
rhodocarpa Hassk., and C. silvestris altera to T. catappa var. 
chlorocarpa Hassk. The form distributed under Robinson PI. 
Rumph. Amb. U1U is exactly Catappa silvestris litorea Rumph. 

QUISQUALIS Linnaeus 

QUISQUALIS INDICA Linn. Sp. PI. ed. 2 (1762) 556 (type). 

Quisqualis pubescens Burm. f. Fl. Ind. (1768) 104 (type!). 
Quis qualis Rumph. Herb. Amb. 5: 71, t. 38. 

This species is not represented in our Amboina collections. 
So far as the original Linnean description shows, the genus 
and the species were based wholly on Rumphius, although he may 
have had botanical material from India or Malaya. The form 
figured by Rumphius is certainly the common and widely dis- 
tributed Malayan and Philippine form that is currently called 
Quisqualis indica Linn. The Linnean reduction has been fol- 



* Retained name, Brussels Congress; Adamaram Adans. (1763) is older. 



MYRTACEAE 



391 



lowed by most authors, but Burman f., Fl. Ind. (1768) 104, 
based his Quisqualis pubescens wholly on the Rumphian Quis 
qualis, and his var. glabra 1. c. t. 28, f. 2, on Javan specimens. 
Quisqualis pubescens Burm. f. is thus an exact synonym of the 
older Q. indica Linn. Poiret, in Lam. Encycl. 6 (1804) 43, 
referred the Rumphian figure to Quisqualis glabra Burm. f., 
which likewise is a synonym of Q. indica Linn. 

MYRTACEAE 

PSIDIUM Linnaeus 

PSIDIUM GUAJAVA Linn. Sp. PL (1753) 470. 

Psidium pomiferum Linn. Sp. PI. ed. 2 (1762) 672. 
Psidium pyriferum Linn. 1. c. 672. 

Psidium cujavus Linn, in Stickman Herb. Amb. (1754) 7, Amoen. 

Acad. 4 (1759) 119 (type!). 
Cujavus domestica Rumph. Herb. Amb. 1 : 140, t. Jf7. 
Cujavus agrestis Rumph. Herb. Amb. 1 : 142, t. ^8. 
Cujavus silvestris Rumph. Herb. Amb. 1: 143. 
Amboina, Robinson PL Rumph. Amb. 202, July 31, 1913, on hills in the 
vicinity of the town of Amboina. 

The three forms described by Rumphius are all apparently 
referable to Psidium guajava Linn., the two forms figured rep- 
resenting the one with the pyriform fruit (Cujavus domestica 
Rumph.), the other with the ovoid or ellipsoid fruit (Cujavus 
agrestis Rumph., the type of Psidium cujavus Linn.). These 
forms have been recognized by some authors as distinct species, 
by others as varieties of Psidium guajava Linn. Both figures 
were originally reduced by Linnaeus, in Stickman Herb. Amb. 
(1754) 7; t. U7 to Psidium guajava Linn, and t. 48 to Psidium 
cujavus Linn., the latter figure being the whole basis of the 
latter species. Following current modern usage both Psidium 
pomiferum Linn, and Psidium pyriferum Linn, are here con- 
sidered as synonyms of Psidium guajava Linn. 

PSIDIUM CUJAVILLUS Burm. f. Fl. Ind. (1768) 114. 
Psidium pumilum Vahl Symb. 2 (1791) 56. 
Psidium angustifolium Lam. Encycl. 3 (1789) 17. 
Cujavillus Rumph. Herb. Amb. 1: 145, t. ^9. 

Amboina, Way tommo, Robinson PI. Rumph. Amb. 203, September 25, 
1913, in waste places at low altitudes, locally known as guayawas china. 

Psidium cujavillus Burm. f. was based primarily on a Javan 
specimen, and Psidium pumilum Vahl on one from Ceylon, 
although both authors cite Cujavillus Rumph. as a synonym, 
and Burman f . took his specific name from Rumphius. By some 



392 RUMPHIUS'S HERBARIUM AMBOINENSE 



authors the species has been erroneously reduced to Psidium 
guajava Linn. Lamarck also cites Cujavillus Rumph. in the 
original description of Psidium angusti folium Lam. Most 
authors who have had occasion to cite the Rumphian figure have 
placed Cujavillus under Psidium pumilum Vahl, but Psidium 
cujavillus Burm. f . is manifestly the same species and is a much 
older name. Pritzel, Ic. Bot. Index, has erroneously listed the 
figure of Cujavillus as Psidium decaspermum Linn. f.=Decas- 
permum fruticosum Forst. ; while Henschel, with equal error, 
placed it under Nelitris jamhosella Gaertn.= Timonius jambo- 
sella Thw. (See also under Decaspermum fruticosum Forst.) 

DECASPERMUM Forster 
DECASPERMUM FRUTICOSUM Forst. Char. Gen. (1776) 74, t. 37. 

Psidium decaspermum Linn. f. Suppl. (1781) 252. 

Eugenia polygama Roxb. Hort. Beng. (1814) 92, nomen nudum, Fl. 

Ind. ed. 2, 2 (1832) 491. 
Nelitris paniculata Lindl. Collect. Bot. (1821) 16. 
Nelitris polygama Spreng. Syst. 2 (1825) 488. 

Nelitris rubra Blume Mus. Bot. 1 (1849) 73, excl. syn. Lour., Poir., 

R. & S. 
Nelitris alba Blume 1. c. 74. 

Decaspermum rubrum Baill. Hist. PI. 6 (1877) 341. 
Decaspermum paniculatum Kurz in Journ. As. Soc. Beng. 46 1 (1877) 
61. 

Caryophyllaster albus Rumph. Herb. Amb. 3: 211. 

Caryophyllaster ruber Rumph. Herb. Amb. 3: 211, t. 136. 
Amboina, Hitoe messen, Gelala, and Batoe gadjah, Robinson PL Rumph. 
Amb. 205, 206, 207, August to September, 1913, in forests and along 
streams, altitude 80 to 250 meters. 

Caryophyllaster ruber Lour, was erroneously reduced by Lou- 
reiro, Fl. Cochinch. (1790) 144, to Antherura rubra Lour., a 
rubiaceous plant, to which the synonyms Psychotria rubra Poir. 
and P. antherura R. & S. pertain, although placed by Blume with 
Caryophyllaster ruber under Nelitris rubra Blume. Poiret, 
Roemer and Schultes, and Blume cite the Rumphian plant as 
a synonym of Psychotria rubra Poir., P. antherura R. & S., and 
Nelitris rubra Blume, respectively. The form described by 
Rumphius as Caryophyllaster albus I consider to be referable 
to the same species as Caryophyllaster ruber, and accordingly 
here reduce it with Nelitris alba Blume to Decaspermum fruti- 
cosum Forst. Decaspermum fruticosum Forst. and Psidium 
decaspermum Linn. f. have been confused with Timonius jam- 
bosella Thw., on account of Gaertner's erroneous reduction of 
Decaspermum fruticosum Forst. to Nelitris jambosella Gaertn. ; 
the species figured by Gaertner is a true Timonius, the Ceylon 



MYRTACEAE 



393 



Timonius jambosella Thw. Gaertner's description is in part 
some species of Eugenia. The plant here considered appears 
in herbaria generally under the name Decaspermum paniculatum 
Kurz, but from material before me I can see no reason for con- 
sidering it specifically distinct from the type of the genus, 
Decaspermum fruticosum Forst. The species is somewhat 
variable. It is a common and widely distributed plant in Malaya 
and Polynesia. 

EUGENIA Linnaeus 

EUGENIA CARYOPHYLLATA Thunb. Diss. (1788) 1. 

Caryophyllus aromaticus Linn. Sp. PI. (1753) 515, non Eugenia aro- 
matica Berg. 

Caryophyllus silvestris Teysm. ex Hassk. in Abh. Naturf. Gesellsch. 

Halle 9 (1866) 167 [Neue Schlussel (1866) 25] (type!). 
Caryophyllum Rumph. Herb. Amb. 2: 1, t. 1. 
Caryophyllum regium Rumph. Herb. Amb. 2: 10, t. 2. 
Caryophyllum silvestre Rumph. Herb. Amb. 2: 12, t. 3. 
Amboina, Kati-kati, Robinson PL Rumph. Amb. 201, October, 1913, from 
cultivated trees, locally known as chenki. 

I can see no valid reason for considering that more than 
one species is represented by the three forms figured and de- 
scribed by Rumphius, although Teysmann, quoted by Hasskarl, 
Neue Schlussel (1866) 25, makes Caryophyllum silvestre Rumph. 
the type of Caryophyllus silvestris Teysm. The first figure 
represents the normal cultivated form, the second a form with 
somewhat fasciated inflorescences, while the third apparently 
represents the wild form of the same species. All three figures 
were originally reduced by Linneaus to Caryophyllus aromaticus 
Linn., in Stickman Herb. Amb. (1754) 8, Amoen. Acad. 4 (1759) 
120, Sp. PL ed. 2 (1762) 735, and this reduction has been very 
generally accepted as the correct disposition of all three. 

EUGENIA AQUEA Burm. f. Fl. Ind. (1768) 114 (type!). 
Jambosa aquea DC. Prodr. 3 (1828) 288 (type!). 
Cerocarpus aqueus Hassk. in Flora 25 (1842) Beibl. 36 (type!). 
Eugenia mindanaensis C. B. Rob. in Philip. Journ. Sci. 4 (1909) 
Bot. 363. 

Jambosa aquea Rumph. Herb. Amb. 1: 126, t. 38, f. 2. 
Jambosa aquea altera Rumph. Herb. Amb. 1 : 126. 

Amboina, Hoenoet, Robinson PL Rumph. Amb. 192, from cultivated trees, 
locally known as jambu ayer. 

Jambosa aquea Rumph. is the whole basis of Eugenia aquea 
Burm. f. and is generally cited in botanical literature under 
Jambosa aquea DC. It was erroneously referred by Lamarck 
to Eugenia javdnica Lam. and by Pritzel to Eugenia racemosa 



. 394 



RUMPHIUS'S HERBARIUM AMBOINENSE 



Linn. Eugenia mindanaensis C. B. Rob., described from Min- 
danao specimens, is manifestly identical with Eugenia aquea 
Burm. f. This reduction had been indicated by Doctor Robin- 
son in the herbarium of the Bureau of Science before his de- 
parture for Amboina in June, 1913. 

EUGENIA CUM IN I (Linn.) comb. nov. 

Myrtus cumini Linn. Sp. PI. (1753) 471. 

Eugenia jambolana Lam. Encycl. 3 (1789) 198. 

Calyptranthes jambolana Willd. in Usteri Ann. 17 (1796) 23. 

Eugenia obtusifolia Roxb. Hort. Beng. (1814) 37, nomen nudum, Fl. 

Ind. ed. 2, 2 (1832) 485. 
Syzygium jambolanum DC. Prodr. 3 (1828) 259. 

Syzygium cumini Skeels in U. S. Dept. Agr. Bur. PI. Ind. Bull. 248 
(1912) 25. 

Jambosa ceramica Rumph. Herb. Amb. 1 : 130, t. hi. 
Jambolana Rumph. Herb. Amb. 1 : 131, t. h2. 

This species is not represented in our Amboina collections, but 
Rel. Robins. 2UUS from Macassar, Celebes, July 11, 1913, is 
typical Jambolana Rumph. Linnaeus, in Stickman Herb. Amb. 
(1745) 7, Amoen. Acad. 4 (1759) 119, reduced Jambolana, 
with doubt, to Jambolifera pedunculata Linn., in which he was 
followed by Burman f., Lamarck, Loureiro, Murray, and Pritzel. 
It has been cited by various authors under Eugenia jambolana 
Lam., Calyptranthes jambolana Willd., Eugenia obtusifolia 
Roxb., and Syzygium jambolanum DC, and is the type of var. 
p of the last in DC. Prodr. 3 (1828) 260. 

I am now of the opinion that Myrtus cumini Linn, supplies 
the oldest valid specific name for this species. Jambolifera 
pedunculata Linn., Sp. PL (1753) 349, was based primarily 
on Fl. Zeyl. 139, and the specimen in Hermann's herbarium 
is Acr onychia laurifolia Blume.* The description in Flora Zey- 
lanica applies to Acr onychia laurifolia, not to Eugenia cumini, 
but the name and the synonyms Jambolones and Jambolons 
apply to Eugenia cumini. It is clear, however, that Linnaeus's 
later conception of Jambolifera pedunculata was as Eugenia 
cumini rather than Acr onychia laurifolia, as shown by his reduc- 
tion of Jambolana Rumph. and the reference to Plukenet added 
in the Mantissa 2 (1771) 371. I believe, however, that the 
original description and specimen should stand as representing 
Jambolifera pedunculata Linn, and that it goes with Acronychia 
laurifolia Blume as a synonym. Myrtus cumini Linn, was based 
wholly on Fl. Zeyl. 185, and the description and the specimen 



* See Trimen in Journ. Linn. Soc. 24 (1887) 140 



MYRTACEAE 



395 



in Hermann's herbarium is Eugenia jambolana Lam. The 
specific names, cumini and jambolifera, were interchanged 
between Eugenia and Act onychia. 

I can see no reason for considering Jambosa ceramica Rumph. 
to be other than Eugenia cumini (Linn.) Merr. Linnaeus placed 
it under Myrtus cumini Linn., and Willdenow placed it under 
Calyptranthes caryophyllifolia Willd,, both synonyms of Eugenia 
cumini (Linn.) Merr. Lamarck, Encycl. 3 (1789) 199, placed 
it under Eugenia cymosa Lam., a species based primarily on 
specimens from the Isle of France. At any rate Jambosa 
ceramica Rumph. is not Eugenia cymosa Lam. as Lamarck's 
species is currently interpreted. 

EUGENIA JAVANICA Lam. Encycl. 3 (1789) 200. 

Jambosa silvestris parvifolia Rumph. Herb. Amb. 1: 129, 2: t. UO. 

In all copies of the Herbarium Amboinense, t. 1+0 of volume 
one and two are transposed ; the plate in volume one corresponds 
to Radix deipariae spuria Rumph. Herb. Amb. 2: 127 and is 
Gmelina villosa Roxb., while the plate in volume two corresponds 
to Jambosa silvestris parvifolia Rumph. Herb. Amb. 1 : 129 and 
is the species here considered (see under Gmelina villosa Roxb., 
p. 454). Perhaps the chief reason why the description and 
figure as given by Rumphius have never been properly placed 
was due to this transposition of the plates, which is mentioned 
in the Auctuarium (Herb. Amb. 7: 3). The illustration seems 
to me to be a fairly typical representation of Eugenia javanica 
Lam. ; and I have made this reduction with considerable confi- 
dence, although I have seen no botanical material from Amboina 
that I would refer to this species. 

EUGENIA SUBGLAUCA Koord. & Valeton in Bull. Inst. Bot. Buitenz. 2 
(1899) 8, var. 
Jambosa litorea Rumph. Herb. Amb. 3: 81, t. 53. 
Amboina, Amahoesoe and Hoenoet, Robinson PI. Rumph. Amb. 193, 196, 
September and October, 1913, on cliffs near the seashore, locally known as 
jambu ayer and jambu puti. 

No previous reduction of Jambosa litorea Rumph. has been 
suggested, other than Henschel's statement that it pertained 
to the Myrtaceae. The description and figure manifestly per- 
tain to Eugenia, a species in the group with Eugenia javanica 
Lam., E. colubcob C. B. Rob., and E. subglauca Koord. & Valeton. 
The Amboina specimens, which certainly represent Jambosa 
litorea Rumph., differ from the Javanese Eugenia subglauca 
Koord. & Val. in their leaves, which are more rounded at the 



396 



RUMPHIUS'S HERBARIUM AMBOINENSE 



base and relatively somewhat broader and thicker. Even if 
eventually distinguished as a valid species, it must certainly be 
placed near the Javanese one. 

EUGENIA RUM PHI I sp. nov. § Syzygium. 

Arbor rubra III Rumph. Herb. Amb. 3: 76. 

Amboina, Hitoe messen, Robinson PL Rumph. Amb. 199 (type), 200, 
November 6, 1913, in forests, altitude about 175 meters, locally known as 
kayu mera. 

Arbor glabra circiter 16 m alta, ramis ramulisque teretibus 
vel ramulis obscure rotundato-angulatis ; foliis obovatis ad 
obovato-ellipticis, usque ad 10 cm longis, apice latissime rotun- 
datis interdum retusis vel obscurissime apiculatis, basi angus- 
tatis, cuneatis, margine revolutis, supra olivaceis, valde nitidis, 
subtus pallidis, haud puncticulatis, nervis utrinque numerosis, 
tenuibus; inflorescentiis corymbosis, terminalibus, circiter 5 cm 
longis; floribus plerumque in triadibus dispositis, omnibus bre- 
viter pedicellatis, calycis circiter 7 mm longis, anguste infundi- 
buliformibus, subtruncatis, calyptra 5 mm diametro. 

An entirely glabrous tree about 16 m in height, the branches 
and branchlets brownish, smooth, terete or sometimes with 
obscure rounded angles. Leaves coriaceous, obovate to obovate- 
elliptic, 5 to 10 cm long, 3 to 5.5 cm wide, apex broadly rounded, 
often retuse, sometimes obscurely apiculate, base gradually nar- 
rowed, cuneate, margins recurved, the upper surface dark-oliva- 
ceous, strongly shining, the lower pale, dull, not puncticulate ; 
lateral nerves numerous, slender, spreading, densely arranged, 
the primary ones but little more prominent than the secondary, 
reticulations obsolete or nearly so; petioles about 1 cm long. 
Inflorescence terminal, corymbose, about 5 cm long and wide, 
branches from the base, the branches ascending, the flowers 
white, mostly in triads on the ultimate branchlets, all pedicelled, 
the bracteoles broadly ovate, rounded, thick, about 1 mm long, 
the pedicels stout, 2 to 3 mm long. Calyx narrowly funnel- 
shaped, about 7 mm long, truncate or with obscure lobes. Ca- 
lyptra 5 mm in diameter, broadly ovoid, rounded. Stamens 
indefinite, 5 to 10 mm long. 

This is a sufficiently characteristic species, which I cannot 
refer to any previously described form. It is readily recogniz- 
able by its densely nerved, coriaceous, obovate to obovate-elliptic, 
broadly rounded, shining leaves, these often retuse at the apex. 
No previous attempt has been made to determine the status of 
Arbor rubra III of Rumphius, other than Hasskarl's reduction 
of it to the Myrtaceae; it is undoubtedly Eugenia rumphii. 



MYRTACEAE 



397 



EUGENIA STIPULARIS (Blume) Miq. Fl. Ind. Bat. 1 1 (1855) 441. 

Gelpkea stipularis Blume Mus. Bot. 1 (1849) 88. 

Jambosa silvestris ayer utan Rumph. Herb. Amb. 1: 129. 
Amboina, Lateri, Robinson PL Rumph. Amb. 191, August 25, 1913, in 
forests, altitude about 200 meters, locally known as jambu utan and kayu 
jambu jambu. 

Hamilton, Mem. Wern. Soc. 5 2 (1826) 338, placed this under 
Eugenia laeta Ham., the description of which, however, was 
based on an Indian specimen. Blume, Mus. Bot. 1 (1849) 104, 
followed Hamilton in this reduction of Jambosa silvestris ayer 
utan and cites Jambosa linearis Korth. as a synonym. The 
description given by Rumphius certainly does not apply to Euge- 
nia laeta Ham. nor to Jambosa linearis Korth., but does apply 
fairly closely to Eugenia stipularis Miq., the type of which was 
from Amboina. 

EUGENIA CELEBICA (Blume) comb. nov. 

Jambosa celebica Blume Mus. Bot. 1 (1849) 107. 
Jambosa silvestris s. biawas Rumph. Herb. Amb. 1: 128? 

The reduction of this Celebes form, very imperfectly described 
by Rumphius, merely follows Blume's suggestion. It may or 
may not be the correct disposition of it, although the form 
that Rumphius described is certainly a Eugenia of the section 
Jambosa. 

EUGENIA JAM BOS Linn. Sp. PI. (1753) 470. 

Jambosa vulgaris DC. Prodr. 3 (1828) 286. 
Jambosa rosacea Rumph. Herb. Amb. 1 : 123. 

The description certainly applies to the rose apple, Eugenia 
jambos Linn. (Jambosa vulgaris DC), where Jambosa rosacea 
Rumph. has been referred by Henschel, Blume, DeVriese, Miquel, 
and Hasskarl. 

EUGENIA sp. aff. jambos Linn.? 

Jambosa (sylvestris) alba Blume Mus. Bot. 1 (1849) 94, non Don, nec 

Eugenia alba Roxb. 
Jambosa silvestris alba Rumph. Herb. Amb. 1: 127, t. 39. 

Rumphius's figure represents a characteristic species, and 
when once collected in Amobina botanical material should 
be readily connected with it. Jambosa silvestris alba Rumph. 
was originally reduced by Linnaeus to Eugenia jambos Linn., 
in Stickman Herb. Amb. (1754) 7, Amoen. Acad. 4 (1759) 119, 
Syst. ed. 10 (1759) 1055, Sp. PL ed. 2 (1762) 672; and in this 
manifestly erroneous reduction he was followed by Burman f., 
Lamarck, Willdenow, Loureiro, Roxburgh, and Persoon. Wight 



398 



RUMPHIUS'S HERBARIUM AMBOINENSE 



and Arnott, Prodr. (1834) 332, with doubt, reduced it to Jambosa 
aquea DC. ; Hasskarl placed it under Eugenia macrophylla DC. ; 
and Berg, erroneously placed it under Jambosa vulgaris DC. 
Blume, Mus. Bot. 1 (1849) 93, has maintained it as a valid 
species, Jambosa alba, but the specific name used by him is 
invalid in both Eugenia and Jambosa. 

EUGENIA MALACCENSIS Linn. Sp. PL (1753) 470. 
Jambosa malaccensis DC. Prodr. 3 (1828) 286. 
Myrtus malaccensis Spreng. Syst. 2 (1825) 484. 
Jambosa purpurascens DC. Prodr. 3 (1828) 286. 

Eugenia purpurea Roxb. Hort. Beng. (1814) 37, nomen nudum, Fl. 

Ind. ed. 2, 2 (1832) 483. 
Jambosa domestica Blume Mus. Bot. 1 (1849) 91. 
Jambosa domestica Rumph. Herb. Amb. 1: 121, t. 37. 
Jambosa nigra Rumph. Herb. Amb. 1: 125, t. 38, f. 1. 

Amboina, Paso, Robinson PI. Rumph. Amb. 19 U, October 29, 1913, from 
cultivated trees, locally known as jambu ruttun. 

Jambosa domestica Rumph. was first reduced to Eugenia ma- 
laccensis Linn, by Linnaeus, Syst. ed. 10 (1759) 1055, and Jam- 
bosa nigra Rumph. was reduced to the same species by Linnaeus, 
in Stickman Herb. Amb. (1754) 7 ; both reductions apparently are 
correct. Neither figure can be considered as good, however. 
The Rumphian names have been cited by various authors under 
one or another of the synonyms given above. The several 
forms named and described by Rumphius were made by Blume, 
Mus. Bot. 1 (1849) 91, the types of several varieties of Jambosa 
domestica B\ume^= Eugenia malaccensis Linn., as follows : 
Jambosa domestica II minor Rumph. =var. minor Blume; Jambosa 
domestica rosacea Rumph. =var. rosacea Blume; Jambosa domestica 
calapparia Rumph. =var. calapparia Blume; and Jambosa nigra 
Rumph. =var. nigra Blume. They are apparently all variants 
of Eugenia malaccensis Linn., which like most cultivated fruit 
trees presents a considerable range of variation in the color, 
size, and other characters of its fruits. 

EUGENIA M ELASTOM I FOLIA (Blume) comb. nov. 

Jambosa melastomif olia Blume Mus. Bot. 1 (1849) 102. 
Arbor rubra II Rumph. Herb. Amb. 3: 76. 

Amboina, Lateri, Robinson PI. Rumph. Amb. 198, September 5, 1913, in 
forests, altitude about 200 meters, locally known as daun jambu jambu. 

The specimen, which presents very young flowers, appears 
to agree with both Rumphius's and with Blume's descriptions, 
Jambosa melastomif olia Blume having been based on Amboina 
specimens collected by Zippel. I do not agree with Miquel, 



MYRTACEAE 



399 



Fl. Ind. Bat. I 1 (1855) 522, in reducing Jambosa melastomi folia 
Blume to Jambosa bifaria Wight. 

EUGENIA sp. 

Arbor rubra I Rumph. Herb. Amb. 3:. 74, t. 47. 

This is a sufficiently characteristic species of Eugenia of 
the section Syzygium; it is, however, not represented in our Am- 
boina collections. No botanist has suggested its determination 
beyond the genus. Lamarck reduced it to Eugenia sp., and 
Teysmann expressed the opinion that it is, with doubt, a Jam- 
bosa. The figure looks suspiciously like Eucalyptus deglupta 
Blume (see p. 401), but although the bark is described as 
peeling off in thin flakes, a character found in many species of 
Eugenia, the description of the fruits and of the odor and taste 
of the leaves definitely removes it from Eucalyptus. 

EUGENIA sp. 

Arbor rubra I angustifolia Rumph. Herb. Amb. 3: 75, t. 48. 

This is a characteristic species of Eugenia of the section 
Jambosa, not represented in our Amboina collections. No sug- 
gestions have been made by other botanists as to its possible or 
probable identity. In the absence of material for comparison 
with the various named species from the Moluccas, no suggestion 
can be made as to its proper disposition. 

EUGENIA sp. 

Folium intinctus Rumph. Herb. Amb. 3: 202. 

Loureiro, Fl. Cochin. (1790) 231, mentions this under Jam- 
bolifera odorata Lour. (=Acronychia?) , with which it has 
nothing in common. Henschel, Vita Rumph. (1833) 160, places 
it, with doubt, under Cyminosma odorata DC, which is merely a 
new name for Jambolifera odorata Lour. Rumphius's descrip- 
tion conforms in all respects to Eugenia and probably refers 
to some species of this genus. 

EUGENIA sp. 

Arbor rubra II saxatilis Rumph. Herb. Amb. 3: 76. 

Further determination of this form is impossible from the data 
and material at present available. 

EUGENIA sp. 

Arbor rubra IV Rumph. Herb. Amb. 3: 77. 

The form described is probably a Eugenia, but its exact 
status is indeterminable from data at present available. 



400 



RUMPHIUS'S HERBARIUM AMBOINENSE 



EUGENIA sp.? 

Perticaria ferrea parvifolia Rumph. Herb. Amb. 3: 80, t. 52. 

The figure resembles Eugenia in many characters, yet if the 
flowers be correctly delineated, the plant can be no Eugenia, 
and hardly a myrtaceous one. Its status is indeterminable from 
material and data at present available. 

EUGENIA sp.? 

Perticaria ferrea latifolia Rumph. Herb. Amb. 3: 80. 

This may or may not belong in the same genus with the 
preceding one. Its status is entirely uncertain, and it may not 
even belong in the Myrtaceae. 

EUGENIA sp. 

Jambosa silvestris alba Rumph. Herb. Amb. 3: 81. 

This is a Eugenia of the section Jambosa, but its further 
identity is entirely doubtful. 

METROSIDEROS * Banks 

M ETROSI DEROS VERA Roxb. Hort. Beng. (1814) 37 (type!); Lindl. 
Collect. Bot. (1821) t. 18. 
Nania vera Miq. Fl. Ind. Bat. 1 1 (1855) 400. 

Syncarpia vertholenii Teysm. & Binn. in Nat. Tijdschr. Ned. Ind. 2 

(1851) 307, Nederl. Kruidk. Arch. 3 (1855) 411. 
Metrosideros vera parvifolia Rumph. Herb. Amb. 3: 16, t. 7. 
Metrosideros vera latifolia Rumph. Herb. Amb. 3: 16, 19. 

This species is not represented in our Amboina collections. 
Metrosideros vera, usually accredited to Lindley as its author, 
should be accredited to Roxburgh instead. The original publica- 
tion of the name was based wholly on Rumphius by citation of the 
illustration, in Hortus Bengalensis (1814) 37.f It was later 
described by Roxburgh, Fl. Ind. ed. 2, 2 (1832) 477, from speci- 
mens cultivated at Calcutta, which had been introduced into the 
botanic garden from Amboina. Syncarpia vertholenii Teysm. 
& Binn. was described from Amboina specimens, and its authors 
reduced to it Metrosideros vera parvifolia Rumph. The first 
reduction was that made by Loureiro, Fl. Cochinch. (1790) 309, 
to Opa metrosideros Lour. Opa metrosideros Lour, is not at 
all Metrosideros vera Roxb., but is a synonym of Raphiolepis 
indica (Linn.) Lindl., of the Rosaceae. The form described by 
Rumphius as Metrosideros vera latifolia does not appear to 

* Retained name, Brussels Congress; Nani Adans. (1763) is older, 
t See C. B. Robinson in Philip. Journ. Sci. 7 (1912) Bot. 414. 



MYRTACEAE 



401 



be specifically distinct from the one described as Metrosideros 
vera par vi folia, although a more extended exploration of the 
Moluccas may show that two distinct species are involved. 

EUCALYPTUS L'Heritier 

EUCALYPTUS DEGLUPTA Blume Mus. Bot. 1 (1849) 83. 
Populus deglubata Reinw. ex Blume 1. c. in syn. 
Eucalyptus versicolor Blume Mus. Bot. 1 (1849) 84 (type!). 
Eucalyptus multiflora Rich ex A. Gray Bot. Wilkes U. S. Explor. 

Exped. (1854) 554. 
Eucalyptus naudiniana F. Muell. in Austral. Journ. Pharm. (1886) 

239, Bot. Centralbl. 28 (1886) 179. 
Eugenia binacag Elm. Lean. Philip. Bot. 7 (1914) 2351. 
Eucalyptus binacag Elm. 1. c. 8 (1915) 2776. 
Arbor versicolor Rumph. Herb. Amb. 3: 122, t. 80. 

Rumphius's material, on which his figure and description of 
Arbor versicolor were based, was from Ceram, not from Am- 
boina. The description and the figure, as far as they go, are 
unmistakably a Eucalyptus. I feel quite confident that Eugenia 
deglupta Blume, from Celebes ; E. versicolor Blume, from Ceram ; 
E. multiflora Rich and E. binacag Elm., of Mindanao; and E. 
naudiniana F. MuelL, of the Bismarck Archipelago, are all refer- 
able to a single species, which is now definitely known from a 
half-dozen localities in Mindanao, from New Guinea, and from 
the Bismarck Archipelago, and with the inclusion of Blume's 
species, from Celebes and Ceram. There is not a character 
given by Blume for either Eucalyptus versicolor or E. deglupta 
by which the two can be definitely distinguished from each 
other or from Eucalyptus naudiniana F.-Muell. Eucalyptus 
moluccana Roxb., as described, must represent a different species, 
at least entirely different from Eucalyptus naudiniana F.-Muell. 
and the Philippine synonyms cited here. Eucalyptus versicolor 
Blume is based wholly on Rumphius's description of Arbor versi- 
color, and it is to be noted that Blume, by error, cites t. 53 instead 
of t. 80 as representing the species. The latter figure is Eugenia 
subglauca Koord. & Valeton, as I have here determined it (see 
p. 395). 

Eucalyptus sarassa Blume, Mus. Bot. 1 (1849) 84, unaccom- 
panied by any word of description, was based on Kaju sarassa 
Rumph., incidentally mentioned by Rumphius, Herb. Amb. 3: 
122, following the description of Arbor versicolor. It is inde- 
terminable from any data now available, and there is little 
or no evidence that it belongs to Eucalyptus. 

144971 26 



402 



RUMPHIUS'S HERBARIUM AMBOINENSE 



LEPTOSPERMUM Forster 

LEPTOSPERMUM FLAVESCENS Smith in Trans. Linn. Soc. 3 (1797) 262. 

Leptospermum amboinense Blume Bijdr. (1826) 1100. 
Leptospermum porophyllum Cav. Ic. 4 (1797) 17, t. 330, f. 2. 
Melaleuca thea Wendl. Sert. Hannov. (1795-98) 24, t. 13? 
Leptospermum thea Willd. Sp. PI. 2 (1799) 949? 

Macklottia amboinensis Korth. in Nederl. Kruidk. Arch. 1 (1847) 196. 
Myrtus amboinensis Rumph. Herb. Amb. 2: 77, t. 18. 
Amboina, Hoetoemoeri road, Robinson PL Rumph. Amb. 20U, September 
30, 1913, on lightly forested hillsides, altitude about 250 meters. 

Myrtus amboinensis Rumph. has been very generally reduced 
to Leptospermum amboinense Blume, but this species does not 
appear to be specifically distinct from the much older Leptosper- 
mum flavescens Smith. The younger Linnaeus reduced it to 
Melaleuca virgata Linn, f., Suppl. (1781) 343; which, however, 
was based on Leptospermum virgatum Forst.=Baeckea virgata 
Andr., in which he was followed by Willdenow and by Lamarck. 
Burman f., Fl. Ind. (1768) 115, had already referred it to Myr- 
tus communis Linn., an entirely erroneous disposition of it. 
Leptospermum thea (Wendl.) Willd. may prove to be the oldest 
valid name for the species. 

MELALEUCA* Linnaeus 

MELALEUCA LEUCADENDRA Linn. Mant. 1 (1767) 105. 

Myrtus leucadendra Linn, in Stickman Herb. Amb. (1754) 9, Amoen. 
Acad. 4 (1759) 120, Syst. ed. 10 (1759) 1056, Sp. PI. ed. 2 (1762) 
676 (type!). 

Myrtus saligna Burm. f. Fl. Ind. (1768) 116, saltern quoad syn. Rumph, 
Melaleuca cajuputi Roxb. Hort. Beng. (1814) 59 (type!). 
Melaleuca saligna Blume Mus. Bot. 1 (1849) 66, non Schauer. 
Melaleuca viridiflora Blume Bijdr. (1026) 1099, var. angustifolia 
Blume 1. c. 

Melaleuca trinervis Ham. ex Henschel Vita Rumph. (1833) 147 
(type!). 

Melaleuca minor Sm. in Rees Cyclop. 23 (1813) no. 2. 
Arbor alba major Rumph. Herb. Amb. 2: 72, t. 16. 
Arbor alba minor Rumph. Herb. Amb. 2: 76, t. 17. 
Amboina, Soja road, Robinson PL Rumph. Amb. 208, August 1, 1913, on 
open hillsides, altitude about 150 meters, locally known as kayu puti. 

Myrtus leucadendra Linn, as originally published included 
both t. 16 and t. 17 of Rumphius. Myrtus leucadendra Linn., 
Melaleuca leucadendra Linn., Melaleuca cajuputi Roxb., M. tri- 
nervis Ham., and possibly M. minor Sm. are all typified by the 
Rumphian descriptions and figures. Some authors have as- 
sumed that t. 17, f. 2, represents a distinct species, which has 



* Retained name, Vienna Code; Cajuputi Adans. (1763) is older. 



MELASTOMATAC